diff --git a/images/sustainable_ai/ai_lifecycle.jpeg b/images/sustainable_ai/ai_lifecycle.jpeg new file mode 100644 index 00000000..0687d89c Binary files /dev/null and b/images/sustainable_ai/ai_lifecycle.jpeg differ diff --git a/images/sustainable_ai/azure_dashboard.png b/images/sustainable_ai/azure_dashboard.png new file mode 100644 index 00000000..e18cee67 Binary files /dev/null and b/images/sustainable_ai/azure_dashboard.png differ diff --git a/images/sustainable_ai/carbon_benchmarks.png b/images/sustainable_ai/carbon_benchmarks.png new file mode 100644 index 00000000..e7970091 Binary files /dev/null and b/images/sustainable_ai/carbon_benchmarks.png differ diff --git a/images/sustainable_ai/energy_datacenter.png b/images/sustainable_ai/energy_datacenter.png new file mode 100644 index 00000000..5dfa54fd Binary files /dev/null and b/images/sustainable_ai/energy_datacenter.png differ diff --git a/images/sustainable_ai/europe_energy_grid.png b/images/sustainable_ai/europe_energy_grid.png new file mode 100644 index 00000000..f89efd45 Binary files /dev/null and b/images/sustainable_ai/europe_energy_grid.png differ diff --git a/images/sustainable_ai/mckinsey_analysis.png b/images/sustainable_ai/mckinsey_analysis.png new file mode 100644 index 00000000..ff36d0d9 Binary files /dev/null and b/images/sustainable_ai/mckinsey_analysis.png differ diff --git a/images/sustainable_ai/model_carbonfootprint.png b/images/sustainable_ai/model_carbonfootprint.png new file mode 100644 index 00000000..11421813 Binary files /dev/null and b/images/sustainable_ai/model_carbonfootprint.png differ diff --git a/images/sustainable_ai/model_scaling.png b/images/sustainable_ai/model_scaling.png new file mode 100644 index 00000000..ef071961 Binary files /dev/null and b/images/sustainable_ai/model_scaling.png differ diff --git a/images/sustainable_ai/statista_chip_growth.png b/images/sustainable_ai/statista_chip_growth.png new file mode 100644 index 00000000..fe007cb1 Binary files /dev/null and b/images/sustainable_ai/statista_chip_growth.png differ diff --git a/references.bib b/references.bib index 1184f00f..dee50899 100644 --- a/references.bib +++ b/references.bib @@ -1,3496 +1,4657 @@ -@article{Ratner_Hancock_Dunnmon_Goldman_Ré_2018, - title = {Snorkel metal: Weak supervision for multi-task learning.}, - author = {Ratner, Alex and Hancock, Braden and Dunnmon, Jared and Goldman, Roger and R\'{e}, Christopher}, - year = 2018, - journal = {Proceedings of the Second Workshop on Data Management for End-To-End Machine Learning}, +@article{/content/paper/876367e3-en, + title = {A blueprint for building national compute capacity for artificial intelligence}, + author = {OECD}, + year = 2023, + number = 350, + doi = {https://doi.org/https://doi.org/10.1787/876367e3-en}, + url = {https://www.oecd-ilibrary.org/content/paper/876367e3-en} } - -@article{oecd22, - author = "OECD", - title = "Measuring the environmental impacts of artificial intelligence compute and applications", - year = "2022", - number = "341", - url = "https://www.oecd-ilibrary.org/content/paper/7babf571-en", - doi = "https://doi.org/https://doi.org/10.1787/7babf571-en" -} - - - -@inproceedings{sculley2015hidden, - title = {"Everyone wants to do the model work, not the data work": Data Cascades in High-Stakes AI}, - author = {Nithya Sambasivan and Shivani Kapania and Hannah Highfill and Diana Akrong and Praveen Kumar Paritosh and Lora Mois Aroyo}, - year = 2021, +@inproceedings{10.1145/1993744.1993791, + title = {Characterizing and Analyzing Renewable Energy Driven Data Centers}, + author = {Li, Chao and Qouneh, Amer and Li, Tao}, + year = 2011, + booktitle = {Proceedings of the ACM SIGMETRICS Joint International Conference on Measurement and Modeling of Computer Systems}, + location = {San Jose, California, USA}, + publisher = {Association for Computing Machinery}, + address = {New York, NY, USA}, + series = {SIGMETRICS '11}, + pages = {131–132}, + doi = {10.1145/1993744.1993791}, + isbn = 9781450308144, + url = {https://doi.org/10.1145/1993744.1993791}, + abstract = {An increasing number of data centers today start to incorporate renewable energy solutions to cap their carbon footprint. However, the impact of renewable energy on large-scale data center design is still not well understood. In this paper, we model and evaluate data centers driven by intermittent renewable energy. Using real-world data center and renewable energy source traces, we show that renewable power utilization and load tuning frequency are two critical metrics for designing sustainable high-performance data centers. Our characterization reveals that load power fluctuation together with the intermittent renewable power supply introduce unnecessary tuning activities, which can increase the management overhead and degrade the performance of renewable energy driven data centers.}, + numpages = 2, + keywords = {data center, power variation, renewable energy, load tuning} } - -@inproceedings{kocher1996timing, - title={Timing attacks on implementations of Diffie-Hellman, RSA, DSS, and other systems}, - author={Kocher, Paul C}, - booktitle={Advances in Cryptology—CRYPTO’96: 16th Annual International Cryptology Conference Santa Barbara, California, USA August 18--22, 1996 Proceedings 16}, - pages={104--113}, - year={1996}, - organization={Springer} -} - -@inproceedings{agrawal2003side, - title={The EM side—channel (s)}, - author={Agrawal, Dakshi and Archambeault, Bruce and Rao, Josyula R and Rohatgi, Pankaj}, - booktitle={Cryptographic Hardware and Embedded Systems-CHES 2002: 4th International Workshop Redwood Shores, CA, USA, August 13--15, 2002 Revised Papers 4}, - pages={29--45}, - year={2003}, - organization={Springer} -} - -@article{breier2018deeplaser, - title={Deeplaser: Practical fault attack on deep neural networks}, - author={Breier, Jakub and Hou, Xiaolu and Jap, Dirmanto and Ma, Lei and Bhasin, Shivam and Liu, Yang}, - journal={arXiv preprint arXiv:1806.05859}, - year={2018} +@article{10242251, + title = {Training Spiking Neural Networks Using Lessons From Deep Learning}, + author = {Eshraghian, Jason K. and Ward, Max and Neftci, Emre O. and Wang, Xinxin and Lenz, Gregor and Dwivedi, Girish and Bennamoun, Mohammed and Jeong, Doo Seok and Lu, Wei D.}, + year = 2023, + journal = {Proceedings of the IEEE}, + volume = 111, + number = 9, + pages = {1016--1054}, + doi = {10.1109/JPROC.2023.3308088}, + bdsk-url-1 = {https://doi.org/10.1109/JPROC.2023.3308088} } - - -@inproceedings{skorobogatov2003optical, - title={Optical fault induction attacks}, - author={Skorobogatov, Sergei P and Anderson, Ross J}, - booktitle={Cryptographic Hardware and Embedded Systems-CHES 2002: 4th International Workshop Redwood Shores, CA, USA, August 13--15, 2002 Revised Papers 4}, - pages={2--12}, - year={2003}, - organization={Springer} -} - -@inproceedings{skorobogatov2009local, - title={Local heating attacks on flash memory devices}, - author={Skorobogatov, Sergei}, - booktitle={2009 IEEE International Workshop on Hardware-Oriented Security and Trust}, - pages={1--6}, - year={2009}, - organization={IEEE} -} - - -@article{oprea2022poisoning, - title={Poisoning Attacks Against Machine Learning: Can Machine Learning Be Trustworthy?}, - author={Oprea, Alina and Singhal, Anoop and Vassilev, Apostol}, - journal={Computer}, - volume={55}, - number={11}, - pages={94--99}, - year={2022}, - publisher={IEEE} -} - -@inproceedings{antonakakis2017understanding, - title={Understanding the mirai botnet}, - author={Antonakakis, Manos and April, Tim and Bailey, Michael and Bernhard, Matt and Bursztein, Elie and Cochran, Jaime and Durumeric, Zakir and Halderman, J Alex and Invernizzi, Luca and Kallitsis, Michalis and others}, - booktitle={26th USENIX security symposium (USENIX Security 17)}, - pages={1093--1110}, - year={2017} +@techreport{7437cc5b-f961-36e9-a745-ba32105a94d0, + title = {The Geopolitics of Critical Minerals Supply Chains}, + author = {Jane Nakano}, + year = 2021, + pages = {II--III}, + url = {http://www.jstor.org/stable/resrep30033.1}, + urldate = {2023-12-08}, + institution = {Center for Strategic and International Studies (CSIS)} } - -@article{goodfellow2020generative, - title={Generative adversarial networks}, - author={Goodfellow, Ian and Pouget-Abadie, Jean and Mirza, Mehdi and Xu, Bing and Warde-Farley, David and Ozair, Sherjil and Courville, Aaron and Bengio, Yoshua}, - journal={Communications of the ACM}, - volume={63}, - number={11}, - pages={139--144}, - year={2020}, - publisher={ACM New York, NY, USA} -} - - -@conference{Rombach22cvpr, -title = {High-Resolution Image Synthesis with Latent Diffusion Models}, -author = {Robin Rombach and Andreas Blattmann and Dominik Lorenz and Patrick Esser and Björn Ommer}, -url = {https://github.com/CompVis/latent-diffusionhttps://arxiv.org/abs/2112.10752}, -year = {2022}, -booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)}, -} - - -@inproceedings{ramesh2021zero, - title={Zero-shot text-to-image generation}, - author={Ramesh, Aditya and Pavlov, Mikhail and Goh, Gabriel and Gray, Scott and Voss, Chelsea and Radford, Alec and Chen, Mark and Sutskever, Ilya}, - booktitle={International Conference on Machine Learning}, - pages={8821--8831}, - year={2021}, - organization={PMLR} -} - -@article{shan2023prompt, - title={Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models}, - author={Shan, Shawn and Ding, Wenxin and Passananti, Josephine and Zheng, Haitao and Zhao, Ben Y}, - journal={arXiv preprint arXiv:2310.13828}, - year={2023} +@article{9563954, + title = {Deep Learning's Diminishing Returns: The Cost of Improvement is Becoming Unsustainable}, + author = {Thompson, Neil C. and Greenewald, Kristjan and Lee, Keeheon and Manso, Gabriel F.}, + year = 2021, + journal = {IEEE Spectrum}, + volume = 58, + number = 10, + pages = {50--55}, + doi = {10.1109/MSPEC.2021.9563954} +} +@article{a4e2073d2c6e4fcb9b7459c2e44716f1, + title = {Electrons have no identity: Setting right misrepresentations in Google and Apple{\textquoteright}s clean energy purchasing}, + author = {Chukwuka Monyei and Kirsten Jenkins}, + year = 2018, + month = dec, + journal = {Energy Research & Social Science}, + publisher = {Elsevier}, + volume = 46, + pages = {48--51}, + doi = {10.1016/j.erss.2018.06.015}, + issn = {2214-6296}, + abstract = {Aside dedicated generation, transmission and distribution networks, the hype around corporations and other entities purchasing so called clean energy may be considered a deliberate accounting misrepresentation. To illustrate this case in this short perspective, we begin by explaining the technical difficulties of remaining “renewables pure”. We then give case studies of two organisations – Apple Inc. and Google LLC – who are, arguably, at fault of making such claims. The method is a simple, non-systematic comparison between what is technically possible, and what is claimed to be possible. Given that incongruous renewables claims have the potential to further impoverish vulnerable households who must bear the financial costs of renewables integration, we conclude that a successful decarbonisation pathway must not have selective winners or losers.}, + keywords = {electricity generation, clean energy, power purchase agreements, renewables}, + language = {English} } - -@article{soufleri2023synthetic, - author = {Efstathia Soufleri and Gobinda Saha and Kaushik Roy}, - date-added = {2023-11-22 19:26:18 -0500}, - date-modified = {2023-11-22 19:26:57 -0500}, - journal = {arXiv preprint arXiv:2210.03205}, - title = {Synthetic Dataset Generation for Privacy-Preserving Machine Learning}, - year = {2023}} - -@article{eldan2023whos, - author = {Ronen Eldan and Mark Russinovich}, - date-added = {2023-11-22 19:24:35 -0500}, - date-modified = {2023-11-22 19:25:20 -0500}, - journal = {arXiv preprint arXiv:2310.02238}, - title = {Who's Harry Potter? Approximate Unlearning in LLMs}, - year = {2023}} - -@article{khan2021knowledgeadaptation, - author = {Mohammad Emtiyaz Khan and Siddharth Swaroop}, - date-added = {2023-11-22 19:22:50 -0500}, - date-modified = {2023-11-22 19:23:40 -0500}, - journal = {arXiv preprint arXiv:2106.08769}, - title = {Knowledge-Adaptation Priors}, - year = {2021}} - -@article{tarun2023deep, - author = {Ayush K Tarun and Vikram S Chundawat and Murari Mandal and Mohan Kankanhalli}, - date-added = {2023-11-22 19:20:59 -0500}, - date-modified = {2023-11-22 19:21:59 -0500}, - journal = {arXiv preprint arXiv:2210.08196}, - title = {Deep Regression Unlearning}, - year = {2023}} - -@article{Li2020Federated, - author = {Li, Tian and Sahu, Anit Kumar and Talwalkar, Ameet and Smith, Virginia}, - date-added = {2023-11-22 19:15:13 -0500}, - date-modified = {2023-11-22 19:17:19 -0500}, - journal = {IEEE Signal Processing Magazine}, - number = {3}, - pages = {50-60}, - title = {Federated Learning: Challenges, Methods, and Future Directions}, - volume = {37}, - year = {2020}} - -@article{MAL-083, - author = {Peter Kairouz and H. Brendan McMahan and Brendan Avent and Aur{\'e}lien Bellet and Mehdi Bennis and Arjun Nitin Bhagoji and Kallista Bonawitz and Zachary Charles and Graham Cormode and Rachel Cummings and Rafael G. L. D'Oliveira and Hubert Eichner and Salim El Rouayheb and David Evans and Josh Gardner and Zachary Garrett and Adri{\`a} Gasc{\'o}n and Badih Ghazi and Phillip B. Gibbons and Marco Gruteser and Zaid Harchaoui and Chaoyang He and Lie He and Zhouyuan Huo and Ben Hutchinson and Justin Hsu and Martin Jaggi and Tara Javidi and Gauri Joshi and Mikhail Khodak and Jakub Konecn{\'y} and Aleksandra Korolova and Farinaz Koushanfar and Sanmi Koyejo and Tancr{\`e}de Lepoint and Yang Liu and Prateek Mittal and Mehryar Mohri and Richard Nock and Ayfer {\"O}zg{\"u}r and Rasmus Pagh and Hang Qi and Daniel Ramage and Ramesh Raskar and Mariana Raykova and Dawn Song and Weikang Song and Sebastian U. Stich and Ziteng Sun and Ananda Theertha Suresh and Florian Tram{\`e}r and Praneeth Vepakomma and Jianyu Wang and Li Xiong and Zheng Xu and Qiang Yang and Felix X. Yu and Han Yu and Sen Zhao}, - date-added = {2023-11-22 19:14:08 -0500}, - date-modified = {2023-11-22 19:14:08 -0500}, - doi = {10.1561/2200000083}, - issn = {1935-8237}, - journal = {Foundations and Trends{\textregistered} in Machine Learning}, - number = {1--2}, - pages = {1-210}, - title = {Advances and Open Problems in Federated Learning}, - url = {http://dx.doi.org/10.1561/2200000083}, - volume = {14}, - year = {2021}, - Bdsk-Url-1 = {http://dx.doi.org/10.1561/2200000083}} - @inproceedings{abadi2016deep, - address = {New York, NY, USA}, - author = {Abadi, Martin and Chu, Andy and Goodfellow, Ian and McMahan, H. Brendan and Mironov, Ilya and Talwar, Kunal and Zhang, Li}, - booktitle = {Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security}, - date-added = {2023-11-22 18:06:03 -0500}, + title = {Deep Learning with Differential Privacy}, + author = {Abadi, Martin and Chu, Andy and Goodfellow, Ian and McMahan, H. Brendan and Mironov, Ilya and Talwar, Kunal and Zhang, Li}, + year = 2016, + booktitle = {Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security}, + publisher = {Association for Computing Machinery}, + address = {New York, NY, USA}, + series = {CCS '16}, + pages = {308--318}, + date-added = {2023-11-22 18:06:03 -0500}, date-modified = {2023-11-22 18:08:42 -0500}, - keywords = {deep learning, differential privacy}, - pages = {308--318}, - publisher = {Association for Computing Machinery}, - series = {CCS '16}, - title = {Deep Learning with Differential Privacy}, - year = {2016}} - -@inproceedings{Dwork2006Theory, - address = {Berlin, Heidelberg}, - author = {Dwork, Cynthia and McSherry, Frank and Nissim, Kobbi and Smith, Adam}, - booktitle = {Theory of Cryptography}, - date-added = {2023-11-22 18:04:12 -0500}, - date-modified = {2023-11-22 18:05:20 -0500}, - editor = {Halevi, Shai and Rabin, Tal}, - pages = {265-284}, - publisher = {Springer Berlin Heidelberg}, - title = {Calibrating Noise to Sensitivity in Private Data Analysis}, - year = {2006}} - -@article{Gupta2023ChatGPT, - author = {Gupta, Maanak and Akiri, Charankumar and Aryal, Kshitiz and Parker, Eli and Praharaj, Lopamudra}, - date-added = {2023-11-22 18:01:41 -0500}, - date-modified = {2023-11-22 18:02:55 -0500}, - journal = {IEEE Access}, - pages = {80218-80245}, - title = {From ChatGPT to ThreatGPT: Impact of Generative AI in Cybersecurity and Privacy}, - volume = {11}, - year = {2023}} - -@inproceedings{Biega2020Oper, - address = {New York, NY, USA}, - author = {Biega, Asia J. and Potash, Peter and Daum\'{e}, Hal and Diaz, Fernando and Finck, Mich\`{e}le}, - booktitle = {Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval}, - date-added = {2023-11-22 17:57:23 -0500}, - date-modified = {2023-11-22 17:59:54 -0500}, - keywords = {data minimization, privacy, gdpr, recommender systems, purpose limitation, personalization}, - pages = {399--408}, - publisher = {Association for Computing Machinery}, - series = {SIGIR '20}, - title = {Operationalizing the Legal Principle of Data Minimization for Personalization}, - year = {2020}} - -@article{cavoukian2009privacy, - author = {Cavoukian, Ann}, - date-added = {2023-11-22 17:55:45 -0500}, - date-modified = {2023-11-22 17:56:58 -0500}, - journal = {Office of the Information and Privacy Commissioner}, - title = {Privacy by design}, - year = {2009}} - -@article{Gao2020Physical, - author = {Gao, Yansong and Al-Sarawi, Said F. and Abbott, Derek}, - date-added = {2023-11-22 17:52:20 -0500}, - date-modified = {2023-11-22 17:54:56 -0500}, - journal = {Nature Electronics}, - month = {February}, - number = {2}, - pages = {81-91}, - title = {Physical unclonable functions}, - volume = {3}, - year = {2020}} - -@inproceedings{Rashmi2018Secure, - author = {R.V. Rashmi and A. Karthikeyan}, - booktitle = {2018 Second International Conference on Electronics, Communication and Aerospace Technology (ICECA)}, - date-added = {2023-11-22 17:50:16 -0500}, - date-modified = {2023-11-22 17:51:39 -0500}, - pages = {291-298}, - title = {Secure boot of Embedded Applications - A Review}, - year = {2018}} - -@article{miller2015remote, - author = {Miller, Charlie and Valasek, Chris}, - date-added = {2023-11-22 17:11:27 -0500}, - date-modified = {2023-11-22 17:12:18 -0500}, - journal = {Black Hat USA}, - number = {S 91}, - pages = {1-91}, - title = {Remote exploitation of an unaltered passenger vehicle}, - volume = {2015}, - year = {2015}} - -@book{dhanjani2015abusing, - author = {Dhanjani, Nitesh}, - date-added = {2023-11-22 17:09:41 -0500}, - date-modified = {2023-11-22 17:10:22 -0500}, - publisher = {O'Reilly Media, Inc.}, - title = {Abusing the internet of things: blackouts, freakouts, and stakeouts}, - year = {2015}} - -@inproceedings{zhao2018fpga, - author = {Zhao, Mark and Suh, G Edward}, - booktitle = {2018 IEEE Symposium on Security and Privacy (SP)}, - date-added = {2023-11-22 17:08:21 -0500}, - date-modified = {2023-11-22 17:09:07 -0500}, - organization = {IEEE}, - pages = {229-244}, - title = {FPGA-based remote power side-channel attacks}, - year = {2018}} - -@inproceedings{gnad2017voltage, - author = {Gnad, Dennis RE and Oboril, Fabian and Tahoori, Mehdi B}, - booktitle = {2017 27th International Conference on Field Programmable Logic and Applications (FPL)}, - date-added = {2023-11-22 17:07:13 -0500}, - date-modified = {2023-11-22 17:07:59 -0500}, - organization = {IEEE}, - pages = {1-7}, - title = {Voltage drop-based fault attacks on FPGAs using valid bitstreams}, - year = {2017}} - -@inproceedings{Asonov2004Keyboard, - author = {Asonov, D. and Agrawal, R.}, - booktitle = {IEEE Symposium on Security and Privacy, 2004. Proceedings. 2004}, - date-added = {2023-11-22 17:05:39 -0500}, - date-modified = {2023-11-22 17:06:45 -0500}, - organization = {IEEE}, - pages = {3-11}, - title = {Keyboard acoustic emanations}, - year = {2004}} - -@article{Burnet1989Spycatcher, - author = {David Burnet and Richard Thomas}, - date-added = {2023-11-22 17:03:00 -0500}, - date-modified = {2023-11-22 17:04:44 -0500}, - journal = {Journal of Law and Society}, - number = {2}, - pages = {210-224}, - title = {Spycatcher: The Commodification of Truth}, - volume = {16}, - year = {1989}} - -@article{Kocher2011Intro, - author = {Kocher, Paul and Jaffe, Joshua and Jun, Benjamin and Rohatgi, Pankaj}, - date-added = {2023-11-22 16:58:42 -0500}, - date-modified = {2023-11-22 17:00:36 -0500}, - journal = {Journal of Cryptographic Engineering}, - month = {April}, - number = {1}, - pages = {5-27}, - title = {Introduction to differential power analysis}, - volume = {1}, - year = {2011}} - -@inproceedings{gandolfi2001electromagnetic, - author = {Gandolfi, Karine and Mourtel, Christophe and Olivier, Francis}, - booktitle = {Cryptographic Hardware and Embedded Systems---CHES 2001: Third International Workshop Paris, France, May 14--16, 2001 Proceedings 3}, - date-added = {2023-11-22 16:56:42 -0500}, - date-modified = {2023-11-22 16:57:40 -0500}, - organization = {Springer}, - pages = {251-261}, - title = {Electromagnetic analysis: Concrete results}, - year = {2001}} - -@inproceedings{kocher1999differential, - author = {Kocher, Paul and Jaffe, Joshua and Jun, Benjamin}, - booktitle = {Advances in Cryptology---CRYPTO'99: 19th Annual International Cryptology Conference Santa Barbara, California, USA, August 15--19, 1999 Proceedings 19}, - date-added = {2023-11-22 16:55:28 -0500}, - date-modified = {2023-11-22 16:56:18 -0500}, - organization = {Springer}, - pages = {388-397}, - title = {Differential power analysis}, - year = {1999}} - -@inproceedings{hsiao2023mavfi, - author = {Hsiao, Yu-Shun and Wan, Zishen and Jia, Tianyu and Ghosal, Radhika and Mahmoud, Abdulrahman and Raychowdhury, Arijit and Brooks, David and Wei, Gu-Yeon and Reddi, Vijay Janapa}, - booktitle = {2023 Design, Automation \& Test in Europe Conference \& Exhibition (DATE)}, - date-added = {2023-11-22 16:54:11 -0500}, - date-modified = {2023-11-22 16:55:12 -0500}, - organization = {IEEE}, - pages = {1-6}, - title = {Mavfi: An end-to-end fault analysis framework with anomaly detection and recovery for micro aerial vehicles}, - year = {2023}} - -@inproceedings{Breier2018Practical, - address = {New York, NY, USA}, - author = {Breier, Jakub and Hou, Xiaolu and Jap, Dirmanto and Ma, Lei and Bhasin, Shivam and Liu, Yang}, - booktitle = {Proceedings of the 2018 ACM SIGSAC Conference on Computer and Communications Security}, - date-added = {2023-11-22 16:51:23 -0500}, - date-modified = {2023-11-22 16:53:46 -0500}, - keywords = {fault attacks, deep learning security, adversarial attacks}, - pages = {2204--2206}, - publisher = {Association for Computing Machinery}, - series = {CCS '18}, - title = {Practical Fault Attack on Deep Neural Networks}} - -@inproceedings{govindavajhala2003using, - author = {Govindavajhala, Sudhakar and Appel, Andrew W}, - booktitle = {2003 Symposium on Security and Privacy, 2003.}, - date-added = {2023-11-22 16:46:13 -0500}, - date-modified = {2023-11-22 16:47:03 -0500}, - organization = {IEEE}, - pages = {154-156}, - title = {Using memory errors to attack a virtual machine}, - year = {2003}} - -@inproceedings{amiel2006fault, - author = {Amiel, Frederic and Clavier, Christophe and Tunstall, Michael}, - booktitle = {International Workshop on Fault Diagnosis and Tolerance in Cryptography}, - date-added = {2023-11-22 16:45:05 -0500}, - date-modified = {2023-11-22 16:45:55 -0500}, - organization = {Springer}, - pages = {223-236}, - title = {Fault analysis of DPA-resistant algorithms}, - year = {2006}} - -@inproceedings{hutter2009contact, - author = {Hutter, Michael and Schmidt, Jorn-Marc and Plos, Thomas}, - booktitle = {2009 European Conference on Circuit Theory and Design}, - date-added = {2023-11-22 16:43:29 -0500}, - date-modified = {2023-11-22 16:44:30 -0500}, - organization = {IEEE}, - pages = {409-412}, - title = {Contact-based fault injections and power analysis on RFID tags}, - year = {2009}} - -@inproceedings{barenghi2010low, - author = {Barenghi, Alessandro and Bertoni, Guido M and Breveglieri, Luca and Pellicioli, Mauro and Pelosi, Gerardo}, - booktitle = {2010 IEEE International Symposium on Hardware-Oriented Security and Trust (HOST)}, - date-added = {2023-11-22 16:42:05 -0500}, - date-modified = {2023-11-22 16:43:09 -0500}, - organization = {IEEE}, - pages = {7-12}, - title = {Low voltage fault attacks to AES}, - year = {2010}} - -@book{joye2012fault, - author = {Joye, Marc and Tunstall, Michael}, - date-added = {2023-11-22 16:35:24 -0500}, - date-modified = {2023-11-22 16:36:20 -0500}, - publisher = {Springer Publishing Company, Incorporated}, - title = {Fault Analysis in Cryptography}, - year = {2012}} - -@inproceedings{Kocher2018spectre, - author = {Paul Kocher and Jann Horn and Anders Fogh and and Daniel Genkin and Daniel Gruss and Werner Haas and Mike Hamburg and Moritz Lipp and Stefan Mangard and Thomas Prescher and Michael Schwarz and Yuval Yarom}, - booktitle = {40th IEEE Symposium on Security and Privacy (S\&P'19)}, - date-added = {2023-11-22 16:33:35 -0500}, - date-modified = {2023-11-22 16:34:01 -0500}, - title = {Spectre Attacks: Exploiting Speculative Execution}, - year = {2019}} - -@inproceedings{Lipp2018meltdown, - author = {Moritz Lipp and Michael Schwarz and Daniel Gruss and Thomas Prescher and Werner Haas and Anders Fogh and Jann Horn and Stefan Mangard and Paul Kocher and Daniel Genkin and Yuval Yarom and Mike Hamburg}, - booktitle = {27th {USENIX} Security Symposium ({USENIX} Security 18)}, - date-added = {2023-11-22 16:32:26 -0500}, - date-modified = {2023-11-22 16:33:08 -0500}, - title = {Meltdown: Reading Kernel Memory from User Space}, - year = {2018}} - -@article{eykholt2018robust, - author = {Kevin Eykholt and Ivan Evtimov and Earlence Fernandes and Bo Li and Amir Rahmati and Chaowei Xiao and Atul Prakash and Tadayoshi Kohno and Dawn Song}, - date-added = {2023-11-22 16:30:51 -0500}, - date-modified = {2023-11-22 16:31:55 -0500}, - journal = {arXiv preprint arXiv:1707.08945}, - title = {Robust Physical-World Attacks on Deep Learning Models}, - year = {2018}} - + keywords = {deep learning, differential privacy} +} +@inproceedings{abadi2016tensorflow, + title = {$\{$TensorFlow$\}$: a system for $\{$Large-Scale$\}$ machine learning}, + author = {Abadi, Mart{\'\i}n and Barham, Paul and Chen, Jianmin and Chen, Zhifeng and Davis, Andy and Dean, Jeffrey and Devin, Matthieu and Ghemawat, Sanjay and Irving, Geoffrey and Isard, Michael and others}, + year = 2016, + booktitle = {12th USENIX symposium on operating systems design and implementation (OSDI 16)}, + pages = {265--283} +} @inproceedings{Abdelkader_2020, - author = {Abdelkader, Ahmed and Curry, Michael J. and Fowl, Liam and Goldstein, Tom and Schwarzschild, Avi and Shu, Manli and Studer, Christoph and Zhu, Chen}, - booktitle = {ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)}, - date-added = {2023-11-22 16:28:31 -0500}, - date-modified = {2023-11-22 16:29:33 -0500}, - title = {Headless Horseman: Adversarial Attacks on Transfer Learning Models}, - year = {2020}} - -@article{parrish2023adversarial, - author = {Alicia Parrish and Hannah Rose Kirk and Jessica Quaye and Charvi Rastogi and Max Bartolo and Oana Inel and Juan Ciro and Rafael Mosquera and Addison Howard and Will Cukierski and D. Sculley and Vijay Janapa Reddi and Lora Aroyo}, - date-added = {2023-11-22 16:24:50 -0500}, - date-modified = {2023-11-22 16:26:30 -0500}, - journal = {arXiv preprint arXiv:2305.14384}, - title = {Adversarial Nibbler: A Data-Centric Challenge for Improving the Safety of Text-to-Image Models}, - year = {2023}} - -@article{hosseini2017deceiving, - author = {Hosseini, Hossein and Kannan, Sreeram and Zhang, Baosen and Poovendran, Radha}, - date-added = {2023-11-22 16:22:18 -0500}, - date-modified = {2023-11-22 16:23:43 -0500}, - journal = {arXiv preprint arXiv:1702.08138}, - title = {Deceiving google's perspective api built for detecting toxic comments}, - year = {2017}} - -@article{biggio2012poisoning, - author = {Biggio, Battista and Nelson, Blaine and Laskov, Pavel}, - date-added = {2023-11-22 16:21:35 -0500}, - date-modified = {2023-11-22 16:22:06 -0500}, - journal = {arXiv preprint arXiv:1206.6389}, - title = {Poisoning attacks against support vector machines}, - year = {2012}} - -@article{oliynyk2023know, - author = {Oliynyk, Daryna and Mayer, Rudolf and Rauber, Andreas}, - date-added = {2023-11-22 16:18:21 -0500}, - date-modified = {2023-11-22 16:20:44 -0500}, - journal = {ACM Comput. Surv.}, - keywords = {model stealing, Machine learning, model extraction}, - month = {July}, - number = {14s}, - title = {I Know What You Trained Last Summer: A Survey on Stealing Machine Learning Models and Defences}, - volume = {55}, - year = {2023}} - -@article{narayanan2006break, - author = {Narayanan, Arvind and Shmatikov, Vitaly}, - date-added = {2023-11-22 16:16:19 -0500}, - date-modified = {2023-11-22 16:16:59 -0500}, - journal = {arXiv preprint cs/0610105}, - title = {How to break anonymity of the netflix prize dataset}, - year = {2006}} - -@article{ateniese2015hacking, - author = {Ateniese, Giuseppe and Mancini, Luigi V and Spognardi, Angelo and Villani, Antonio and Vitali, Domenico and Felici, Giovanni}, - date-added = {2023-11-22 16:14:42 -0500}, - date-modified = {2023-11-22 16:15:42 -0500}, - journal = {International Journal of Security and Networks}, - number = {3}, - pages = {137-150}, - title = {Hacking smart machines with smarter ones: How to extract meaningful data from machine learning classifiers}, - volume = {10}, - year = {2015}} - -@article{miller2019lessons, - author = {Miller, Charlie}, - date-added = {2023-11-22 16:12:04 -0500}, - date-modified = {2023-11-22 16:13:31 -0500}, - journal = {IEEE Design & Test}, - number = {6}, - pages = {7-9}, - title = {Lessons learned from hacking a car}, - volume = {36}, - year = {2019}} - -@article{farwell2011stuxnet, - author = {Farwell, James P and Rohozinski, Rafal}, - date-added = {2023-11-22 14:03:31 -0500}, - date-modified = {2023-11-22 14:05:19 -0500}, - journal = {Survival}, - number = {1}, - pages = {23-40}, - title = {Stuxnet and the future of cyber war}, - volume = {53}, - year = {2011}} - -@inproceedings{krishnan2023archgym, - author = {Krishnan, Srivatsan and Yazdanbakhsh, Amir and Prakash, Shvetank and Jabbour, Jason and Uchendu, Ikechukwu and Ghosh, Susobhan and Boroujerdian, Behzad and Richins, Daniel and Tripathy, Devashree and Faust, Aleksandra and Janapa Reddi, Vijay}, - booktitle = {Proceedings of the 50th Annual International Symposium on Computer Architecture}, - pages = {1--16}, - title = {ArchGym: An Open-Source Gymnasium for Machine Learning Assisted Architecture Design}, - year = {2023}} - -@misc{kuzmin2022fp8, + title = {Headless Horseman: Adversarial Attacks on Transfer Learning Models}, + author = {Abdelkader, Ahmed and Curry, Michael J. and Fowl, Liam and Goldstein, Tom and Schwarzschild, Avi and Shu, Manli and Studer, Christoph and Zhu, Chen}, + year = 2020, + booktitle = {ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)}, + date-added = {2023-11-22 16:28:31 -0500}, + date-modified = {2023-11-22 16:29:33 -0500} +} +@article{adagrad, + title = {Adaptive Subgradient Methods for Online Learning and Stochastic Optimization}, + author = {John Duchi and Elad Hazan and Yoram Singer}, + year = 2011, + journal = {Journal of Machine Learning Research}, + url = {http://jmlr.org/papers/v12/duchi11a.html} +} +@misc{adam, + title = {Adam: A Method for Stochastic Optimization}, + author = {Diederik P. Kingma and Jimmy Ba}, + year = 2017, + booktitle = {3rd International Conference on Learning Representations, {ICLR} 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings}, + url = {http://arxiv.org/abs/1412.6980}, + eprint = {1412.6980}, archiveprefix = {arXiv}, - author = {Andrey Kuzmin and Mart Van Baalen and Yuwei Ren and Markus Nagel and Jorn Peters and Tijmen Blankevoort}, - eprint = {2208.09225}, primaryclass = {cs.LG}, - title = {FP8 Quantization: The Power of the Exponent}, - year = {2022}} - -@inproceedings{abadi2016tensorflow, - author = {Abadi, Mart{\'\i}n and Barham, Paul and Chen, Jianmin and Chen, Zhifeng and Davis, Andy and Dean, Jeffrey and Devin, Matthieu and Ghemawat, Sanjay and Irving, Geoffrey and Isard, Michael and others}, - booktitle = {12th USENIX symposium on operating systems design and implementation (OSDI 16)}, - pages = {265--283}, - title = {$\{$TensorFlow$\}$: a system for $\{$Large-Scale$\}$ machine learning}, - year = 2016} - -@article{shastri2021photonics, - author = {Shastri, Bhavin J and Tait, Alexander N and Ferreira de Lima, Thomas and Pernice, Wolfram HP and Bhaskaran, Harish and Wright, C David and Prucnal, Paul R}, - journal = {Nature Photonics}, - number = {2}, - pages = {102--114}, - publisher = {Nature Publishing Group UK London}, - title = {Photonics for artificial intelligence and neuromorphic computing}, - volume = {15}, - year = {2021}} - -@inproceedings{jouppi2017datacenter, - author = {Jouppi, Norman P and Young, Cliff and Patil, Nishant and Patterson, David and Agrawal, Gaurav and Bajwa, Raminder and Bates, Sarah and Bhatia, Suresh and Boden, Nan and Borchers, Al and others}, - booktitle = {Proceedings of the 44th annual international symposium on computer architecture}, - pages = {1--12}, - title = {In-datacenter performance analysis of a tensor processing unit}, - year = {2017}} - -@inproceedings{ignatov2018ai, - author = {Ignatov, Andrey and Timofte, Radu and Chou, William and Wang, Ke and Wu, Max and Hartley, Tim and Van Gool, Luc}, - booktitle = {Proceedings of the European Conference on Computer Vision (ECCV) Workshops}, - pages = {0--0}, - title = {Ai benchmark: Running deep neural networks on android smartphones}, - year = {2018}} - + editor = {Yoshua Bengio and Yann LeCun}, + timestamp = {Thu, 25 Jul 2019 14:25:37 +0200}, + biburl = {https://dblp.org/rec/journals/corr/KingmaB14.bib}, + bibsource = {dblp computer science bibliography, https://dblp.org} +} +@misc{adelta, + title = {ADADELTA: An Adaptive Learning Rate Method}, + author = {Matthew D. Zeiler}, + year = 2012, + eprint = {1212.5701}, + archiveprefix = {arXiv}, + primaryclass = {cs.LG} +} @inproceedings{adolf2016fathom, - author = {Adolf, Robert and Rama, Saketh and Reagen, Brandon and Wei, Gu-Yeon and Brooks, David}, - booktitle = {2016 IEEE International Symposium on Workload Characterization (IISWC)}, - organization = {IEEE}, - pages = {1--10}, - title = {Fathom: Reference workloads for modern deep learning methods}, - year = 2016} - + title = {Fathom: Reference workloads for modern deep learning methods}, + author = {Adolf, Robert and Rama, Saketh and Reagen, Brandon and Wei, Gu-Yeon and Brooks, David}, + year = 2016, + booktitle = {2016 IEEE International Symposium on Workload Characterization (IISWC)}, + pages = {1--10}, + organization = {IEEE} +} +@article{afib, + title = {Mobile Photoplethysmographic Technology to Detect Atrial Fibrillation}, + author = {Yutao Guo and Hao Wang and Hui Zhang and Tong Liu and Zhaoguang Liang and Yunlong Xia and Li Yan and Yunli Xing and Haili Shi and Shuyan Li and Yanxia Liu and Fan Liu and Mei Feng and Yundai Chen and Gregory Y.H. Lip and null null}, + year = 2019, + journal = {Journal of the American College of Cardiology}, + volume = 74, + number = 19, + pages = {2365--2375}, + doi = {10.1016/j.jacc.2019.08.019}, + bdsk-url-1 = {https://doi.org/10.1016/j.jacc.2019.08.019} +} +@inproceedings{agarwal2018reductions, + title = {A reductions approach to fair classification}, + author = {Agarwal, Alekh and Beygelzimer, Alina and Dud{\'\i}k, Miroslav and Langford, John and Wallach, Hanna}, + year = 2018, + booktitle = {International conference on machine learning}, + pages = {60--69}, + organization = {PMLR} +} +@inproceedings{agrawal2003side, + title = {The EM side—channel (s)}, + author = {Agrawal, Dakshi and Archambeault, Bruce and Rao, Josyula R and Rohatgi, Pankaj}, + year = 2003, + booktitle = {Cryptographic Hardware and Embedded Systems-CHES 2002: 4th International Workshop Redwood Shores, CA, USA, August 13--15, 2002 Revised Papers 4}, + pages = {29--45}, + organization = {Springer} +} +@article{ai_health_rise, + title = {The rise of Artificial Intelligence in healthcare applications}, + author = {Bohr, Adam and Memarzadeh, Kaveh}, + year = 2020, + month = {Jun}, + journal = {Artificial Intelligence in Healthcare}, + pages = {25–60}, + doi = {10.1016/b978-0-12-818438-7.00002-2} +} @misc{al2016theano, + title = {Theano: A Python framework for fast computation of mathematical expressions}, + author = {The Theano Development Team and Rami Al-Rfou and Guillaume Alain and Amjad Almahairi and Christof Angermueller and Dzmitry Bahdanau and Nicolas Ballas and Fr{\'e}d{\'e}ric Bastien and Justin Bayer and Anatoly Belikov and Alexander Belopolsky and Yoshua Bengio and Arnaud Bergeron and James Bergstra and Valentin Bisson and Josh Bleecher Snyder and Nicolas Bouchard and Nicolas Boulanger-Lewandowski and Xavier Bouthillier and Alexandre de Br{\'e}bisson and Olivier Breuleux and Pierre-Luc Carrier and Kyunghyun Cho and Jan Chorowski and Paul Christiano and Tim Cooijmans and Marc-Alexandre C{\^o}t{\'e} and Myriam C{\^o}t{\'e} and Aaron Courville and Yann N. Dauphin and Olivier Delalleau and Julien Demouth and Guillaume Desjardins and Sander Dieleman and Laurent Dinh and M{\'e}lanie Ducoffe and Vincent Dumoulin and Samira Ebrahimi Kahou and Dumitru Erhan and Ziye Fan and Orhan Firat and Mathieu Germain and Xavier Glorot and Ian Goodfellow and Matt Graham and Caglar Gulcehre and Philippe Hamel and Iban Harlouchet and Jean-Philippe Heng and Bal{\'a}zs Hidasi and Sina Honari and Arjun Jain and S{\'e}bastien Jean and Kai Jia and Mikhail Korobov and Vivek Kulkarni and Alex Lamb and Pascal Lamblin and Eric Larsen and C{\'e}sar Laurent and Sean Lee and Simon Lefrancois and Simon Lemieux and Nicholas L{\'e}onard and Zhouhan Lin and Jesse A. Livezey and Cory Lorenz and Jeremiah Lowin and Qianli Ma and Pierre-Antoine Manzagol and Olivier Mastropietro and Robert T. McGibbon and Roland Memisevic and Bart van Merri{\"e}nboer and Vincent Michalski and Mehdi Mirza and Alberto Orlandi and Christopher Pal and Razvan Pascanu and Mohammad Pezeshki and Colin Raffel and Daniel Renshaw and Matthew Rocklin and Adriana Romero and Markus Roth and Peter Sadowski and John Salvatier and Fran{\c c}ois Savard and Jan Schl{\"u}ter and John Schulman and Gabriel Schwartz and Iulian Vlad Serban and Dmitriy Serdyuk and Samira Shabanian and {\'E}tienne Simon and Sigurd Spieckermann and S. Ramana Subramanyam and Jakub Sygnowski and J{\'e}r{\'e}mie Tanguay and Gijs van Tulder and Joseph Turian and Sebastian Urban and Pascal Vincent and Francesco Visin and Harm de Vries and David Warde-Farley and Dustin J. Webb and Matthew Willson and Kelvin Xu and Lijun Xue and Li Yao and Saizheng Zhang and Ying Zhang}, + year = 2016, archiveprefix = {arXiv}, - author = {The Theano Development Team and Rami Al-Rfou and Guillaume Alain and Amjad Almahairi and Christof Angermueller and Dzmitry Bahdanau and Nicolas Ballas and Fr{\'e}d{\'e}ric Bastien and Justin Bayer and Anatoly Belikov and Alexander Belopolsky and Yoshua Bengio and Arnaud Bergeron and James Bergstra and Valentin Bisson and Josh Bleecher Snyder and Nicolas Bouchard and Nicolas Boulanger-Lewandowski and Xavier Bouthillier and Alexandre de Br{\'e}bisson and Olivier Breuleux and Pierre-Luc Carrier and Kyunghyun Cho and Jan Chorowski and Paul Christiano and Tim Cooijmans and Marc-Alexandre C{\^o}t{\'e} and Myriam C{\^o}t{\'e} and Aaron Courville and Yann N. Dauphin and Olivier Delalleau and Julien Demouth and Guillaume Desjardins and Sander Dieleman and Laurent Dinh and M{\'e}lanie Ducoffe and Vincent Dumoulin and Samira Ebrahimi Kahou and Dumitru Erhan and Ziye Fan and Orhan Firat and Mathieu Germain and Xavier Glorot and Ian Goodfellow and Matt Graham and Caglar Gulcehre and Philippe Hamel and Iban Harlouchet and Jean-Philippe Heng and Bal{\'a}zs Hidasi and Sina Honari and Arjun Jain and S{\'e}bastien Jean and Kai Jia and Mikhail Korobov and Vivek Kulkarni and Alex Lamb and Pascal Lamblin and Eric Larsen and C{\'e}sar Laurent and Sean Lee and Simon Lefrancois and Simon Lemieux and Nicholas L{\'e}onard and Zhouhan Lin and Jesse A. Livezey and Cory Lorenz and Jeremiah Lowin and Qianli Ma and Pierre-Antoine Manzagol and Olivier Mastropietro and Robert T. McGibbon and Roland Memisevic and Bart van Merri{\"e}nboer and Vincent Michalski and Mehdi Mirza and Alberto Orlandi and Christopher Pal and Razvan Pascanu and Mohammad Pezeshki and Colin Raffel and Daniel Renshaw and Matthew Rocklin and Adriana Romero and Markus Roth and Peter Sadowski and John Salvatier and Fran{\c c}ois Savard and Jan Schl{\"u}ter and John Schulman and Gabriel Schwartz and Iulian Vlad Serban and Dmitriy Serdyuk and Samira Shabanian and {\'E}tienne Simon and Sigurd Spieckermann and S. Ramana Subramanyam and Jakub Sygnowski and J{\'e}r{\'e}mie Tanguay and Gijs van Tulder and Joseph Turian and Sebastian Urban and Pascal Vincent and Francesco Visin and Harm de Vries and David Warde-Farley and Dustin J. Webb and Matthew Willson and Kelvin Xu and Lijun Xue and Li Yao and Saizheng Zhang and Ying Zhang}, - eprint = {1605.02688}, - primaryclass = {cs.SC}, - title = {Theano: A Python framework for fast computation of mathematical expressions}, - year = 2016} - + eprint = {1605.02688}, + primaryclass = {cs.SC} +} @article{Aledhari_Razzak_Parizi_Saeed_2020, - author = {Aledhari, Mohammed and Razzak, Rehma and Parizi, Reza M. and Saeed, Fahad}, - doi = {10.1109/access.2020.3013541}, - journal = {IEEE Access}, - pages = {140699--140725}, - title = {Federated learning: A survey on enabling technologies, Protocols, and applications}, - volume = 8, - year = 2020, - Bdsk-Url-1 = {https://doi.org/10.1109/access.2020.3013541}} - + title = {Federated learning: A survey on enabling technologies, Protocols, and applications}, + author = {Aledhari, Mohammed and Razzak, Rehma and Parizi, Reza M. and Saeed, Fahad}, + year = 2020, + journal = {IEEE Access}, + volume = 8, + pages = {140699--140725}, + doi = {10.1109/access.2020.3013541}, + bdsk-url-1 = {https://doi.org/10.1109/access.2020.3013541} +} +@article{alghamdi2022beyond, + title = {Beyond Adult and COMPAS: Fair multi-class prediction via information projection}, + author = {Alghamdi, Wael and Hsu, Hsiang and Jeong, Haewon and Wang, Hao and Michalak, Peter and Asoodeh, Shahab and Calmon, Flavio}, + year = 2022, + journal = {Advances in Neural Information Processing Systems}, + volume = 35, + pages = {38747--38760} +} @article{aljundi_gradient_nodate, - author = {Aljundi, Rahaf and Lin, Min and Goujaud, Baptiste and Bengio, Yoshua}, - file = {Aljundi et al. - Gradient based sample selection for online continu.pdf:/Users/alex/Zotero/storage/GPHM4KY7/Aljundi et al. - Gradient based sample selection for online continu.pdf:application/pdf}, - language = {en}, - title = {Gradient based sample selection for online continual learning}} - + title = {Gradient based sample selection for online continual learning}, + author = {Aljundi, Rahaf and Lin, Min and Goujaud, Baptiste and Bengio, Yoshua}, + file = {Aljundi et al. - Gradient based sample selection for online continu.pdf:/Users/alex/Zotero/storage/GPHM4KY7/Aljundi et al. - Gradient based sample selection for online continu.pdf:application/pdf}, + language = {en} +} +@online{alleghenycounty2018, + title = {Allegheny Family Screening Tool: Algorithmic Risk Assessment in Child Protective Services}, + author = {Allegheny County Department of Human Services}, + year = 2018, + url = {https://www.alleghenycountyanalytics.us/wp-content/uploads/2018/10/17-ACDHS-11_AFST_102518.pdf}, + organization = {Allegheny County Analytics} +} @inproceedings{altayeb2022classifying, - author = {Altayeb, Moez and Zennaro, Marco and Rovai, Marcelo}, - booktitle = {Proceedings of the 2022 ACM Conference on Information Technology for Social Good}, - pages = {132--137}, - title = {Classifying mosquito wingbeat sound using TinyML}, - year = 2022} - + title = {Classifying mosquito wingbeat sound using TinyML}, + author = {Altayeb, Moez and Zennaro, Marco and Rovai, Marcelo}, + year = 2022, + booktitle = {Proceedings of the 2022 ACM Conference on Information Technology for Social Good}, + pages = {132--137} +} +@inproceedings{amiel2006fault, + title = {Fault analysis of DPA-resistant algorithms}, + author = {Amiel, Frederic and Clavier, Christophe and Tunstall, Michael}, + year = 2006, + booktitle = {International Workshop on Fault Diagnosis and Tolerance in Cryptography}, + pages = {223--236}, + date-added = {2023-11-22 16:45:05 -0500}, + date-modified = {2023-11-22 16:45:55 -0500}, + organization = {Springer} +} @misc{amodei_ai_2018, - author = {Amodei, Dario and Hernandez, Danny}, - journal = {OpenAI Blog}, - month = may, - title = {{AI} and {Compute}}, - url = {https://openai.com/research/ai-and-compute}, - year = 2018, - Bdsk-Url-1 = {https://openai.com/research/ai-and-compute}} - + title = {{AI} and {Compute}}, + author = {Amodei, Dario and Hernandez, Danny}, + year = 2018, + month = may, + journal = {OpenAI Blog}, + url = {https://openai.com/research/ai-and-compute}, + bdsk-url-1 = {https://openai.com/research/ai-and-compute} +} +@article{amodei2016concrete, + title = {Concrete problems in AI safety}, + author = {Amodei, Dario and Olah, Chris and Steinhardt, Jacob and Christiano, Paul and Schulman, John and Man{\'e}, Dan}, + year = 2016, + journal = {arXiv preprint arXiv:1606.06565} +} +@misc{anthony2020carbontracker, + author = {Lasse F. Wolff Anthony and Benjamin Kanding and Raghavendra Selvan}, + year = 2020, + month = {July}, + note = {arXiv:2007.03051}, + howpublished = {ICML Workshop on Challenges in Deploying and monitoring Machine Learning Systems} +} @inproceedings{antol2015vqa, - author = {Antol, Stanislaw and Agrawal, Aishwarya and Lu, Jiasen and Mitchell, Margaret and Batra, Dhruv and Zitnick, C Lawrence and Parikh, Devi}, - booktitle = {Proceedings of the IEEE international conference on computer vision}, - pages = {2425--2433}, - title = {Vqa: Visual question answering}, - year = 2015} - + title = {Vqa: Visual question answering}, + author = {Antol, Stanislaw and Agrawal, Aishwarya and Lu, Jiasen and Mitchell, Margaret and Batra, Dhruv and Zitnick, C Lawrence and Parikh, Devi}, + year = 2015, + booktitle = {Proceedings of the IEEE international conference on computer vision}, + pages = {2425--2433} +} +@inproceedings{antonakakis2017understanding, + title = {Understanding the mirai botnet}, + author = {Antonakakis, Manos and April, Tim and Bailey, Michael and Bernhard, Matt and Bursztein, Elie and Cochran, Jaime and Durumeric, Zakir and Halderman, J Alex and Invernizzi, Luca and Kallitsis, Michalis and others}, + year = 2017, + booktitle = {26th USENIX security symposium (USENIX Security 17)}, + pages = {1093--1110} +} @article{app112211073, + title = {Hardware/Software Co-Design for TinyML Voice-Recognition Application on Resource Frugal Edge Devices}, + author = {Kwon, Jisu and Park, Daejin}, + year = 2021, + journal = {Applied Sciences}, + volume = 11, + number = 22, + doi = {10.3390/app112211073}, + issn = {2076-3417}, + url = {https://www.mdpi.com/2076-3417/11/22/11073}, article-number = 11073, - author = {Kwon, Jisu and Park, Daejin}, - doi = {10.3390/app112211073}, - issn = {2076-3417}, - journal = {Applied Sciences}, - number = 22, - title = {Hardware/Software Co-Design for TinyML Voice-Recognition Application on Resource Frugal Edge Devices}, - url = {https://www.mdpi.com/2076-3417/11/22/11073}, - volume = 11, - year = 2021, - Bdsk-Url-1 = {https://www.mdpi.com/2076-3417/11/22/11073}, - Bdsk-Url-2 = {https://doi.org/10.3390/app112211073}} - + bdsk-url-1 = {https://www.mdpi.com/2076-3417/11/22/11073}, + bdsk-url-2 = {https://doi.org/10.3390/app112211073} +} @article{Ardila_Branson_Davis_Henretty_Kohler_Meyer_Morais_Saunders_Tyers_Weber_2020, - author = {Ardila, Rosana and Branson, Megan and Davis, Kelly and Henretty, Michael and Kohler, Michael and Meyer, Josh and Morais, Reuben and Saunders, Lindsay and Tyers, Francis M. and Weber, Gregor}, - journal = {Proceedings of the 12th Conference on Language Resources and Evaluation}, - month = {May}, - pages = {4218-4222}, - title = {Common Voice: A Massively-Multilingual Speech Corpus}, - year = 2020} - + title = {Common Voice: A Massively-Multilingual Speech Corpus}, + author = {Ardila, Rosana and Branson, Megan and Davis, Kelly and Henretty, Michael and Kohler, Michael and Meyer, Josh and Morais, Reuben and Saunders, Lindsay and Tyers, Francis M. and Weber, Gregor}, + year = 2020, + month = {May}, + journal = {Proceedings of the 12th Conference on Language Resources and Evaluation}, + pages = {4218--4222} +} +@inproceedings{Asonov2004Keyboard, + title = {Keyboard acoustic emanations}, + author = {Asonov, D. and Agrawal, R.}, + year = 2004, + booktitle = {IEEE Symposium on Security and Privacy, 2004. Proceedings. 2004}, + pages = {3--11}, + date-added = {2023-11-22 17:05:39 -0500}, + date-modified = {2023-11-22 17:06:45 -0500}, + organization = {IEEE} +} +@article{ateniese2015hacking, + title = {Hacking smart machines with smarter ones: How to extract meaningful data from machine learning classifiers}, + author = {Ateniese, Giuseppe and Mancini, Luigi V and Spognardi, Angelo and Villani, Antonio and Vitali, Domenico and Felici, Giovanni}, + year = 2015, + journal = {International Journal of Security and Networks}, + volume = 10, + number = 3, + pages = {137--150}, + date-added = {2023-11-22 16:14:42 -0500}, + date-modified = {2023-11-22 16:15:42 -0500} +} +@inproceedings{athalye2018obfuscated, + title = {Obfuscated gradients give a false sense of security: Circumventing defenses to adversarial examples}, + author = {Athalye, Anish and Carlini, Nicholas and Wagner, David}, + year = 2018, + booktitle = {International conference on machine learning}, + pages = {274--283}, + organization = {PMLR} +} @misc{awq, - author = {Lin and Tang, Tang and Yang, Dang and Gan, Han}, - doi = {10.48550/arXiv.2306.00978}, - title = {AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration}, - url = {https://arxiv.org/abs/2306.00978}, - urldate = {2023-10-03}, - year = 2023, - Bdsk-Url-1 = {https://arxiv.org/abs/2306.00978}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.2306.00978}} - + title = {AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration}, + author = {Lin and Tang, Tang and Yang, Dang and Gan, Han}, + year = 2023, + journal = {arXiv}, + doi = {10.48550/arXiv.2306.00978}, + url = {https://arxiv.org/abs/2306.00978}, + urldate = {2023-10-03}, + bdsk-url-1 = {https://arxiv.org/abs/2306.00978}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.2306.00978} +} +@misc{bailey_enabling_2018, + title = {Enabling {Cheaper} {Design}}, + author = {Bailey, Brian}, + year = 2018, + month = sep, + journal = {Semiconductor Engineering}, + url = {https://semiengineering.com/enabling-cheaper-design/}, + urldate = {2023-11-07}, + abstract = {Enabling Cheaper Design, At what point does cheaper design enable a significant growth in custom semiconductor content? Not everyone is onboard with the idea.}, + language = {en-US}, + bdsk-url-1 = {https://semiengineering.com/enabling-cheaper-design/} +} +@article{bains2020business, + title = {The business of building brains}, + author = {Bains, Sunny}, + year = 2020, + journal = {Nat. Electron}, + volume = 3, + number = 7, + pages = {348--351} +} @inproceedings{bamoumen2022tinyml, - author = {Bamoumen, Hatim and Temouden, Anas and Benamar, Nabil and Chtouki, Yousra}, - booktitle = {2022 International Conference on Innovation and Intelligence for Informatics, Computing, and Technologies (3ICT)}, - organization = {IEEE}, - pages = {338--343}, - title = {How TinyML Can be Leveraged to Solve Environmental Problems: A Survey}, - year = 2022} - + title = {How TinyML Can be Leveraged to Solve Environmental Problems: A Survey}, + author = {Bamoumen, Hatim and Temouden, Anas and Benamar, Nabil and Chtouki, Yousra}, + year = 2022, + booktitle = {2022 International Conference on Innovation and Intelligence for Informatics, Computing, and Technologies (3ICT)}, + pages = {338--343}, + organization = {IEEE} +} @article{banbury2020benchmarking, - author = {Banbury, Colby R and Reddi, Vijay Janapa and Lam, Max and Fu, William and Fazel, Amin and Holleman, Jeremy and Huang, Xinyuan and Hurtado, Robert and Kanter, David and Lokhmotov, Anton and others}, - journal = {arXiv preprint arXiv:2003.04821}, - title = {Benchmarking tinyml systems: Challenges and direction}, - year = 2020} - + title = {Benchmarking tinyml systems: Challenges and direction}, + author = {Banbury, Colby R and Reddi, Vijay Janapa and Lam, Max and Fu, William and Fazel, Amin and Holleman, Jeremy and Huang, Xinyuan and Hurtado, Robert and Kanter, David and Lokhmotov, Anton and others}, + year = 2020, + journal = {arXiv preprint arXiv:2003.04821} +} @article{bank2023autoencoders, - author = {Bank, Dor and Koenigstein, Noam and Giryes, Raja}, - journal = {Machine Learning for Data Science Handbook: Data Mining and Knowledge Discovery Handbook}, - pages = {353--374}, - publisher = {Springer}, - title = {Autoencoders}, - year = 2023} - + title = {Autoencoders}, + author = {Bank, Dor and Koenigstein, Noam and Giryes, Raja}, + year = 2023, + journal = {Machine Learning for Data Science Handbook: Data Mining and Knowledge Discovery Handbook}, + publisher = {Springer}, + pages = {353--374} +} +@inproceedings{barenghi2010low, + title = {Low voltage fault attacks to AES}, + author = {Barenghi, Alessandro and Bertoni, Guido M and Breveglieri, Luca and Pellicioli, Mauro and Pelosi, Gerardo}, + year = 2010, + booktitle = {2010 IEEE International Symposium on Hardware-Oriented Security and Trust (HOST)}, + pages = {7--12}, + date-added = {2023-11-22 16:42:05 -0500}, + date-modified = {2023-11-22 16:43:09 -0500}, + organization = {IEEE} +} @book{barroso2019datacenter, - author = {Barroso, Luiz Andr{\'e} and H{\"o}lzle, Urs and Ranganathan, Parthasarathy}, - publisher = {Springer Nature}, - title = {The datacenter as a computer: Designing warehouse-scale machines}, - year = 2019} - + title = {The datacenter as a computer: Designing warehouse-scale machines}, + author = {Barroso, Luiz Andr{\'e} and H{\"o}lzle, Urs and Ranganathan, Parthasarathy}, + year = 2019, + publisher = {Springer Nature} +} +@techreport{bashmakov2022climate, + title = {Climate Change 2022: Mitigation of Climate Change. Contribution of Working Group III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change, Chapter 11}, + author = {Bashmakov, IA and Nilsson, LJ and Acquaye, A and Bataille, C and Cullen, JM and Fischedick, M and Geng, Y and Tanaka, K}, + year = 2022, + institution = {Lawrence Berkeley National Lab.(LBNL), Berkeley, CA (United States)} +} +@inproceedings{bau2017network, + title = {Network dissection: Quantifying interpretability of deep visual representations}, + author = {Bau, David and Zhou, Bolei and Khosla, Aditya and Oliva, Aude and Torralba, Antonio}, + year = 2017, + booktitle = {Proceedings of the IEEE conference on computer vision and pattern recognition}, + pages = {6541--6549} +} +@misc{bayes_hyperparam, + title = {Practical Bayesian Optimization of Machine Learning Algorithms}, + author = {Jasper Snoek and Hugo Larochelle and Ryan P. Adams}, + year = 2012, + eprint = {1206.2944}, + archiveprefix = {arXiv}, + primaryclass = {stat.ML} +} +@article{beck1998beyond, + title = {Beyond linearity by default: Generalized additive models}, + author = {Beck, Nathaniel and Jackman, Simon}, + year = 1998, + journal = {American Journal of Political Science}, + publisher = {JSTOR}, + pages = {596--627} +} @article{Bender_Friedman_2018, - author = {Bender, Emily M. and Friedman, Batya}, - doi = {10.1162/tacl_a_00041}, - journal = {Transactions of the Association for Computational Linguistics}, - pages = {587-604}, - title = {Data statements for natural language processing: Toward mitigating system bias and enabling better science}, - volume = 6, - year = 2018, - Bdsk-Url-1 = {https://doi.org/10.1162/tacl_a_00041}} - + title = {Data statements for natural language processing: Toward mitigating system bias and enabling better science}, + author = {Bender, Emily M. and Friedman, Batya}, + year = 2018, + journal = {Transactions of the Association for Computational Linguistics}, + volume = 6, + pages = {587--604}, + doi = {10.1162/tacl_a_00041}, + bdsk-url-1 = {https://doi.org/10.1162/tacl_a_00041} +} @article{beyer2020we, - author = {Beyer, Lucas and H{\'e}naff, Olivier J and Kolesnikov, Alexander and Zhai, Xiaohua and Oord, A{\"a}ron van den}, - journal = {arXiv preprint arXiv:2006.07159}, - title = {Are we done with imagenet?}, - year = 2020} - + title = {Are we done with imagenet?}, + author = {Beyer, Lucas and H{\'e}naff, Olivier J and Kolesnikov, Alexander and Zhai, Xiaohua and Oord, A{\"a}ron van den}, + year = 2020, + journal = {arXiv preprint arXiv:2006.07159} +} +@inproceedings{bhagoji2018practical, + title = {Practical black-box attacks on deep neural networks using efficient query mechanisms}, + author = {Bhagoji, Arjun Nitin and He, Warren and Li, Bo and Song, Dawn}, + year = 2018, + booktitle = {Proceedings of the European conference on computer vision (ECCV)}, + pages = {154--169} +} +@inproceedings{Biega2020Oper, + title = {Operationalizing the Legal Principle of Data Minimization for Personalization}, + author = {Biega, Asia J. and Potash, Peter and Daum\'{e}, Hal and Diaz, Fernando and Finck, Mich\`{e}le}, + year = 2020, + booktitle = {Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval}, + publisher = {Association for Computing Machinery}, + address = {New York, NY, USA}, + series = {SIGIR '20}, + pages = {399--408}, + date-added = {2023-11-22 17:57:23 -0500}, + date-modified = {2023-11-22 17:59:54 -0500}, + keywords = {data minimization, privacy, gdpr, recommender systems, purpose limitation, personalization} +} +@misc{bigbatch, + title = {ImageNet Training in Minutes}, + author = {Yang You and Zhao Zhang and Cho-Jui Hsieh and James Demmel and Kurt Keutzer}, + year = 2018, + eprint = {1709.05011}, + archiveprefix = {arXiv}, + primaryclass = {cs.CV} +} +@article{biggio2012poisoning, + title = {Poisoning attacks against support vector machines}, + author = {Biggio, Battista and Nelson, Blaine and Laskov, Pavel}, + year = 2012, + journal = {arXiv preprint arXiv:1206.6389}, + date-added = {2023-11-22 16:21:35 -0500}, + date-modified = {2023-11-22 16:22:06 -0500} +} @article{biggio2014pattern, - author = {Biggio, Battista and Fumera, Giorgio and Roli, Fabio}, - journal = {International Journal of Pattern Recognition and Artificial Intelligence}, - number = {07}, - pages = 1460002, - publisher = {World Scientific}, - title = {Pattern recognition systems under attack: Design issues and research challenges}, - volume = 28, - year = 2014} - + title = {Pattern recognition systems under attack: Design issues and research challenges}, + author = {Biggio, Battista and Fumera, Giorgio and Roli, Fabio}, + year = 2014, + journal = {International Journal of Pattern Recognition and Artificial Intelligence}, + publisher = {World Scientific}, + volume = 28, + number = {07}, + pages = 1460002 +} +@article{biggs2021natively, + title = {A natively flexible 32-bit Arm microprocessor}, + author = {Biggs, John and Myers, James and Kufel, Jedrzej and Ozer, Emre and Craske, Simon and Sou, Antony and Ramsdale, Catherine and Williamson, Ken and Price, Richard and White, Scott}, + year = 2021, + journal = {Nature}, + publisher = {Nature Publishing Group UK London}, + volume = 595, + number = 7868, + pages = {532--536} +} +@article{binkert2011gem5, + title = {The gem5 simulator}, + author = {Binkert, Nathan and Beckmann, Bradford and Black, Gabriel and Reinhardt, Steven K and Saidi, Ali and Basu, Arkaprava and Hestness, Joel and Hower, Derek R and Krishna, Tushar and Sardashti, Somayeh and others}, + year = 2011, + journal = {ACM SIGARCH computer architecture news}, + publisher = {ACM New York, NY, USA}, + volume = 39, + number = 2, + pages = {1--7} +} @misc{blalock_what_2020, - abstract = {Neural network pruning---the task of reducing the size of a network by removing parameters---has been the subject of a great deal of work in recent years. We provide a meta-analysis of the literature, including an overview of approaches to pruning and consistent findings in the literature. After aggregating results across 81 papers and pruning hundreds of models in controlled conditions, our clearest finding is that the community suffers from a lack of standardized benchmarks and metrics. This deficiency is substantial enough that it is hard to compare pruning techniques to one another or determine how much progress the field has made over the past three decades. To address this situation, we identify issues with current practices, suggest concrete remedies, and introduce ShrinkBench, an open-source framework to facilitate standardized evaluations of pruning methods. We use ShrinkBench to compare various pruning techniques and show that its comprehensive evaluation can prevent common pitfalls when comparing pruning methods.}, - author = {Blalock, Davis and Ortiz, Jose Javier Gonzalez and Frankle, Jonathan and Guttag, John}, - doi = {10.48550/arXiv.2003.03033}, - file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/MA4QGZ6E/Blalock et al. - 2020 - What is the State of Neural Network Pruning.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/8DFKG4GL/2003.html:text/html}, - keywords = {Computer Science - Machine Learning, Statistics - Machine Learning}, - month = mar, - note = {arXiv:2003.03033 [cs, stat]}, - publisher = {arXiv}, - title = {What is the {State} of {Neural} {Network} {Pruning}?}, - url = {http://arxiv.org/abs/2003.03033}, - urldate = {2023-10-20}, - year = 2020, - Bdsk-Url-1 = {http://arxiv.org/abs/2003.03033}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.2003.03033}} - + title = {What is the {State} of {Neural} {Network} {Pruning}?}, + author = {Blalock, Davis and Ortiz, Jose Javier Gonzalez and Frankle, Jonathan and Guttag, John}, + year = 2020, + month = mar, + publisher = {arXiv}, + doi = {10.48550/arXiv.2003.03033}, + url = {http://arxiv.org/abs/2003.03033}, + urldate = {2023-10-20}, + note = {arXiv:2003.03033 [cs, stat]}, + abstract = {Neural network pruning---the task of reducing the size of a network by removing parameters---has been the subject of a great deal of work in recent years. We provide a meta-analysis of the literature, including an overview of approaches to pruning and consistent findings in the literature. After aggregating results across 81 papers and pruning hundreds of models in controlled conditions, our clearest finding is that the community suffers from a lack of standardized benchmarks and metrics. This deficiency is substantial enough that it is hard to compare pruning techniques to one another or determine how much progress the field has made over the past three decades. To address this situation, we identify issues with current practices, suggest concrete remedies, and introduce ShrinkBench, an open-source framework to facilitate standardized evaluations of pruning methods. We use ShrinkBench to compare various pruning techniques and show that its comprehensive evaluation can prevent common pitfalls when comparing pruning methods.}, + file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/MA4QGZ6E/Blalock et al. - 2020 - What is the State of Neural Network Pruning.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/8DFKG4GL/2003.html:text/html}, + keywords = {Computer Science - Machine Learning, Statistics - Machine Learning}, + bdsk-url-1 = {http://arxiv.org/abs/2003.03033}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.2003.03033} +} +@inproceedings{bondi2018spot, + title = {Spot poachers in action: Augmenting conservation drones with automatic detection in near real time}, + author = {Bondi, Elizabeth and Fang, Fei and Hamilton, Mark and Kar, Debarun and Dmello, Donnabell and Choi, Jongmoo and Hannaford, Robert and Iyer, Arvind and Joppa, Lucas and Tambe, Milind and others}, + year = 2018, + booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence}, + volume = 32, + number = 1 +} +@inproceedings{bourtoule2021machine, + title = {Machine unlearning}, + author = {Bourtoule, Lucas and Chandrasekaran, Varun and Choquette-Choo, Christopher A and Jia, Hengrui and Travers, Adelin and Zhang, Baiwu and Lie, David and Papernot, Nicolas}, + year = 2021, + booktitle = {2021 IEEE Symposium on Security and Privacy (SP)}, + pages = {141--159}, + organization = {IEEE} +} +@article{breier2018deeplaser, + title = {Deeplaser: Practical fault attack on deep neural networks}, + author = {Breier, Jakub and Hou, Xiaolu and Jap, Dirmanto and Ma, Lei and Bhasin, Shivam and Liu, Yang}, + year = 2018, + journal = {arXiv preprint arXiv:1806.05859} +} +@inproceedings{Breier2018Practical, + title = {Practical Fault Attack on Deep Neural Networks}, + author = {Breier, Jakub and Hou, Xiaolu and Jap, Dirmanto and Ma, Lei and Bhasin, Shivam and Liu, Yang}, + booktitle = {Proceedings of the 2018 ACM SIGSAC Conference on Computer and Communications Security}, + publisher = {Association for Computing Machinery}, + address = {New York, NY, USA}, + series = {CCS '18}, + pages = {2204--2206}, + date-added = {2023-11-22 16:51:23 -0500}, + date-modified = {2023-11-22 16:53:46 -0500}, + keywords = {fault attacks, deep learning security, adversarial attacks} +} +@article{bricken2023towards, + title = {Towards Monosemanticity: Decomposing Language Models With Dictionary Learning}, + author = {Bricken, Trenton and Templeton, Adly and Batson, Joshua and Chen, Brian and Jermyn, Adam and Conerly, Tom and Turner, Nick and Anil, Cem and Denison, Carson and Askell, Amanda and others}, + year = 2023, + journal = {Transformer Circuits Thread} +} +@inproceedings{brown_language_2020, + title = {Language {Models} are {Few}-{Shot} {Learners}}, + author = {Brown, Tom and Mann, Benjamin and Ryder, Nick and Subbiah, Melanie and Kaplan, Jared D and Dhariwal, Prafulla and Neelakantan, Arvind and Shyam, Pranav and Sastry, Girish and Askell, Amanda and Agarwal, Sandhini and Herbert-Voss, Ariel and Krueger, Gretchen and Henighan, Tom and Child, Rewon and Ramesh, Aditya and Ziegler, Daniel and Wu, Jeffrey and Winter, Clemens and Hesse, Chris and Chen, Mark and Sigler, Eric and Litwin, Mateusz and Gray, Scott and Chess, Benjamin and Clark, Jack and Berner, Christopher and McCandlish, Sam and Radford, Alec and Sutskever, Ilya and Amodei, Dario}, + year = 2020, + journal = {Advances in neural information processing systems}, + booktitle = {Advances in {Neural} {Information} {Processing} {Systems}}, + publisher = {Curran Associates, Inc.}, + volume = 33, + pages = {1877--1901}, + url = {https://proceedings.neurips.cc/paper_files/paper/2020/hash/1457c0d6bfcb4967418bfb8ac142f64a-Abstract.html}, + urldate = {2023-11-07}, + abstract = {We demonstrate that scaling up language models greatly improves task-agnostic, few-shot performance, sometimes even becoming competitive with prior state-of-the-art fine-tuning approaches. Specifically, we train GPT-3, an autoregressive language model with 175 billion parameters, 10x more than any previous non-sparse language model, and test its performance in the few-shot setting. For all tasks, GPT-3 is applied without any gradient updates or fine-tuning, with tasks and few-shot demonstrations specified purely via text interaction with the model. GPT-3 achieves strong performance on many NLP datasets, including translation, question-answering, and cloze tasks. We also identify some datasets where GPT-3's few-shot learning still struggles, as well as some datasets where GPT-3 faces methodological issues related to training on large web corpora.}, + bdsk-url-1 = {https://proceedings.neurips.cc/paper_files/paper/2020/hash/1457c0d6bfcb4967418bfb8ac142f64a-Abstract.html} +} @article{brown2020language, - author = {Brown, Tom and Mann, Benjamin and Ryder, Nick and Subbiah, Melanie and Kaplan, Jared D and Dhariwal, Prafulla and Neelakantan, Arvind and Shyam, Pranav and Sastry, Girish and Askell, Amanda and others}, - journal = {Advances in neural information processing systems}, - pages = {1877--1901}, - title = {Language models are few-shot learners}, - volume = 33, - year = 2020} - + title = {Language models are few-shot learners}, + author = {Brown, Tom and Mann, Benjamin and Ryder, Nick and Subbiah, Melanie and Kaplan, Jared D and Dhariwal, Prafulla and Neelakantan, Arvind and Shyam, Pranav and Sastry, Girish and Askell, Amanda and others}, + year = 2020, + journal = {Advances in neural information processing systems}, + volume = 33, + pages = {1877--1901} +} +@inproceedings{buolamwini2018genderShades, + title = {Gender shades: Intersectional accuracy disparities in commercial gender classification}, + author = {Buolamwini, Joy and Gebru, Timnit}, + year = 2018, + booktitle = {Conference on fairness, accountability and transparency}, + pages = {77--91}, + organization = {PMLR} +} +@article{Burnet1989Spycatcher, + title = {Spycatcher: The Commodification of Truth}, + author = {David Burnet and Richard Thomas}, + year = 1989, + journal = {Journal of Law and Society}, + volume = 16, + number = 2, + pages = {210--224}, + date-added = {2023-11-22 17:03:00 -0500}, + date-modified = {2023-11-22 17:04:44 -0500} +} +@article{burr2016recent, + title = {Recent progress in phase-change memory technology}, + author = {Burr, Geoffrey W and Brightsky, Matthew J and Sebastian, Abu and Cheng, Huai-Yu and Wu, Jau-Yi and Kim, Sangbum and Sosa, Norma E and Papandreou, Nikolaos and Lung, Hsiang-Lan and Pozidis, Haralampos and others}, + year = 2016, + journal = {IEEE Journal on Emerging and Selected Topics in Circuits and Systems}, + publisher = {IEEE}, + volume = 6, + number = 2, + pages = {146--162} +} +@misc{buyya2010energyefficient, + title = {Energy-Efficient Management of Data Center Resources for Cloud Computing: A Vision, Architectural Elements, and Open Challenges}, + author = {Rajkumar Buyya and Anton Beloglazov and Jemal Abawajy}, + year = 2010, + eprint = {1006.0308}, + archiveprefix = {arXiv}, + primaryclass = {cs.DC} +} @inproceedings{cai_online_2021, - address = {Montreal, QC, Canada}, - author = {Cai, Zhipeng and Sener, Ozan and Koltun, Vladlen}, - booktitle = {2021 {IEEE}/{CVF} {International} {Conference} on {Computer} {Vision} ({ICCV})}, - doi = {10.1109/ICCV48922.2021.00817}, - file = {Cai et al. - 2021 - Online Continual Learning with Natural Distributio.pdf:/Users/alex/Zotero/storage/R7ZMIM4K/Cai et al. - 2021 - Online Continual Learning with Natural Distributio.pdf:application/pdf}, - isbn = {978-1-66542-812-5}, - language = {en}, - month = oct, - pages = {8261--8270}, - publisher = {IEEE}, - shorttitle = {Online {Continual} {Learning} with {Natural} {Distribution} {Shifts}}, - title = {Online {Continual} {Learning} with {Natural} {Distribution} {Shifts}: {An} {Empirical} {Study} with {Visual} {Data}}, - url = {https://ieeexplore.ieee.org/document/9710740/}, - urldate = {2023-10-26}, - year = 2021, - Bdsk-Url-1 = {https://ieeexplore.ieee.org/document/9710740/}, - Bdsk-Url-2 = {https://doi.org/10.1109/ICCV48922.2021.00817}} - + title = {Online {Continual} {Learning} with {Natural} {Distribution} {Shifts}: {An} {Empirical} {Study} with {Visual} {Data}}, + shorttitle = {Online {Continual} {Learning} with {Natural} {Distribution} {Shifts}}, + author = {Cai, Zhipeng and Sener, Ozan and Koltun, Vladlen}, + year = 2021, + month = oct, + booktitle = {2021 {IEEE}/{CVF} {International} {Conference} on {Computer} {Vision} ({ICCV})}, + publisher = {IEEE}, + address = {Montreal, QC, Canada}, + pages = {8261--8270}, + doi = {10.1109/ICCV48922.2021.00817}, + isbn = {978-1-66542-812-5}, + url = {https://ieeexplore.ieee.org/document/9710740/}, + urldate = {2023-10-26}, + file = {Cai et al. - 2021 - Online Continual Learning with Natural Distributio.pdf:/Users/alex/Zotero/storage/R7ZMIM4K/Cai et al. - 2021 - Online Continual Learning with Natural Distributio.pdf:application/pdf}, + language = {en}, + bdsk-url-1 = {https://ieeexplore.ieee.org/document/9710740/}, + bdsk-url-2 = {https://doi.org/10.1109/ICCV48922.2021.00817} +} @article{cai_tinytl_nodate, - author = {Cai, Han and Gan, Chuang and Zhu, Ligeng and Han, Song}, - file = {Cai et al. - TinyTL Reduce Memory, Not Parameters for Efficient.pdf:/Users/alex/Zotero/storage/J9C8PTCX/Cai et al. - TinyTL Reduce Memory, Not Parameters for Efficient.pdf:application/pdf}, - language = {en}, - title = {{TinyTL}: {Reduce} {Memory}, {Not} {Parameters} for {Efficient} {On}-{Device} {Learning}}} - + title = {{TinyTL}: {Reduce} {Memory}, {Not} {Parameters} for {Efficient} {On}-{Device} {Learning}}, + author = {Cai, Han and Gan, Chuang and Zhu, Ligeng and Han, Song}, + year = 2020, + journal = {Advances in Neural Information Processing Systems}, + volume = 33, + pages = {11285--11297}, + file = {Cai et al. - TinyTL Reduce Memory, Not Parameters for Efficient.pdf:/Users/alex/Zotero/storage/J9C8PTCX/Cai et al. - TinyTL Reduce Memory, Not Parameters for Efficient.pdf:application/pdf}, + language = {en} +} @article{cai2018proxylessnas, - author = {Cai, Han and Zhu, Ligeng and Han, Song}, - journal = {arXiv preprint arXiv:1812.00332}, - title = {Proxylessnas: Direct neural architecture search on target task and hardware}, - year = 2018} - + title = {Proxylessnas: Direct neural architecture search on target task and hardware}, + author = {Cai, Han and Zhu, Ligeng and Han, Song}, + year = 2018, + journal = {arXiv preprint arXiv:1812.00332} +} @article{cai2020tinytl, - author = {Cai, Han and Gan, Chuang and Zhu, Ligeng and Han, Song}, - journal = {Advances in Neural Information Processing Systems}, - pages = {11285--11297}, - title = {Tinytl: Reduce memory, not parameters for efficient on-device learning}, - volume = 33, - year = 2020} - + title = {Tinytl: Reduce memory, not parameters for efficient on-device learning}, + author = {Cai, Han and Gan, Chuang and Zhu, Ligeng and Han, Song}, + year = 2020, + journal = {Advances in Neural Information Processing Systems}, + volume = 33, + pages = {11285--11297} +} +@article{calvo2020supporting, + title = {Supporting human autonomy in AI systems: A framework for ethical enquiry}, + author = {Calvo, Rafael A and Peters, Dorian and Vold, Karina and Ryan, Richard M}, + year = 2020, + journal = {Ethics of digital well-being: A multidisciplinary approach}, + publisher = {Springer}, + pages = {31--54} +} +@article{Carbon_LNN, + title = {Carbon emissions and large neural network training}, + author = {Patterson, David and Gonzalez, Joseph and Le, Quoc and Liang, Chen and Munguia, Lluis-Miquel and Rothchild, Daniel and So, David and Texier, Maud and Dean, Jeff}, + year = 2021, + journal = {arXiv preprint arXiv:2104.10350} +} +@inproceedings{carlini2016hidden, + title = {Hidden voice commands}, + author = {Carlini, Nicholas and Mishra, Pratyush and Vaidya, Tavish and Zhang, Yuankai and Sherr, Micah and Shields, Clay and Wagner, David and Zhou, Wenchao}, + year = 2016, + booktitle = {25th USENIX security symposium (USENIX security 16)}, + pages = {513--530} +} +@inproceedings{carlini2017adversarial, + title = {Adversarial examples are not easily detected: Bypassing ten detection methods}, + author = {Carlini, Nicholas and Wagner, David}, + year = 2017, + booktitle = {Proceedings of the 10th ACM workshop on artificial intelligence and security}, + pages = {3--14} +} +@inproceedings{carlini2023extracting, + title = {Extracting training data from diffusion models}, + author = {Carlini, Nicolas and Hayes, Jamie and Nasr, Milad and Jagielski, Matthew and Sehwag, Vikash and Tramer, Florian and Balle, Borja and Ippolito, Daphne and Wallace, Eric}, + year = 2023, + booktitle = {32nd USENIX Security Symposium (USENIX Security 23)}, + pages = {5253--5270} +} +@article{cavoukian2009privacy, + title = {Privacy by design}, + author = {Cavoukian, Ann}, + year = 2009, + journal = {Office of the Information and Privacy Commissioner}, + date-added = {2023-11-22 17:55:45 -0500}, + date-modified = {2023-11-22 17:56:58 -0500} +} +@article{cenci_eco-friendly_2022, + title = {Eco-{Friendly} {Electronics}—{A} {Comprehensive} {Review}}, + author = {Cenci, Marcelo Pilotto and Scarazzato, Tatiana and Munchen, Daniel Dotto and Dartora, Paula Cristina and Veit, Hugo Marcelo and Bernardes, Andrea Moura and Dias, Pablo R.}, + year = 2022, + journal = {Advanced Materials Technologies}, + volume = 7, + number = 2, + pages = 2001263, + doi = {https://doi.org/10.1002/admt.202001263}, + url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/admt.202001263}, + note = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/admt.202001263}, + abstract = {Abstract Eco-friendliness is becoming an indispensable feature for electrical and electronic equipment to thrive in the competitive market. This comprehensive review is the first to define eco-friendly electronics in its multiple meanings: power saving devices, end-of-life impact attenuators, equipment whose manufacturing uses green processing, electronics that use materials that minimize environmental and health risks, designs that improve lifespan, reparability, etc. More specifically, this review discusses eco-friendly technologies and materials that are being introduced to replace the well-established ones. This is done for all material classes (metals, polymers, ceramics, and composites). Manufacturing, recycling, and final product characteristics are discussed in their various interconnected aspects. Additionally, the concept of consciously planned obsolescence is introduced to address the paradoxical relationship between durability and efficiency. The overall conclusions are that there is an important global trend to make electronics more eco-friendly. However, matching the performance and stability of well-established materials and technologies seems to be the main barrier to achieve it. These new implementations can have detrimental or beneficial net impacts on the environment. Assessing their net outcome is challenging because their impacts are frequently unknown and the current evaluation methods (and tools) are incapable of comprehensively quantifying these impacts and generating reliable verdicts.}, + keywords = {eco-friendly electronics, end-of-life, green electronics, green manufacturing, ICT sustainability, recycling, sustainable materials} +} @article{Chapelle_Scholkopf_Zien, - author = {Chapelle, O. and Scholkopf, B. and Zien, Eds., A.}, - doi = {10.1109/tnn.2009.2015974}, - journal = {IEEE Transactions on Neural Networks}, - number = 3, - pages = {542--542}, - title = {Semi-supervised learning (Chapelle, O. et al., eds.; 2006) [book reviews]}, - volume = 20, - year = 2009, - Bdsk-Url-1 = {https://doi.org/10.1109/tnn.2009.2015974}} - + title = {Semi-supervised learning (Chapelle, O. et al., eds.; 2006) [book reviews]}, + author = {Chapelle, O. and Scholkopf, B. and Zien, Eds., A.}, + year = 2009, + journal = {IEEE Transactions on Neural Networks}, + volume = 20, + number = 3, + pages = {542--542}, + doi = {10.1109/tnn.2009.2015974}, + bdsk-url-1 = {https://doi.org/10.1109/tnn.2009.2015974} +} @misc{chen__inpainting_2022, - abstract = {Some simple examples for showing how to use tensor decomposition to reconstruct fluid dynamics}, - author = {Chen (陈新宇), Xinyu}, - journal = {Medium}, - language = {en}, - month = mar, - title = {Inpainting {Fluid} {Dynamics} with {Tensor} {Decomposition} ({NumPy})}, - url = {https://medium.com/@xinyu.chen/inpainting-fluid-dynamics-with-tensor-decomposition-numpy-d84065fead4d}, - urldate = {2023-10-20}, - year = 2022, - Bdsk-Url-1 = {https://medium.com/@xinyu.chen/inpainting-fluid-dynamics-with-tensor-decomposition-numpy-d84065fead4d}} - + title = {Inpainting {Fluid} {Dynamics} with {Tensor} {Decomposition} ({NumPy})}, + author = {Chen (陈新宇), Xinyu}, + year = 2022, + month = mar, + journal = {Medium}, + url = {https://medium.com/@xinyu.chen/inpainting-fluid-dynamics-with-tensor-decomposition-numpy-d84065fead4d}, + urldate = {2023-10-20}, + abstract = {Some simple examples for showing how to use tensor decomposition to reconstruct fluid dynamics}, + language = {en}, + bdsk-url-1 = {https://medium.com/@xinyu.chen/inpainting-fluid-dynamics-with-tensor-decomposition-numpy-d84065fead4d} +} @misc{chen_tvm_2018, - annote = {Comment: Significantly improved version, add automated optimization}, - author = {Chen, Tianqi and Moreau, Thierry and Jiang, Ziheng and Zheng, Lianmin and Yan, Eddie and Cowan, Meghan and Shen, Haichen and Wang, Leyuan and Hu, Yuwei and Ceze, Luis and Guestrin, Carlos and Krishnamurthy, Arvind}, - file = {Chen et al. - 2018 - TVM An Automated End-to-End Optimizing Compiler f.pdf:/Users/alex/Zotero/storage/QR8MHJ38/Chen et al. - 2018 - TVM An Automated End-to-End Optimizing Compiler f.pdf:application/pdf}, - keywords = {Computer Science - Artificial Intelligence, Computer Science - Machine Learning, Computer Science - Programming Languages}, - language = {en}, - month = oct, - note = {arXiv:1802.04799 [cs]}, - publisher = {arXiv}, - shorttitle = {{TVM}}, - title = {{TVM}: {An} {Automated} {End}-to-{End} {Optimizing} {Compiler} for {Deep} {Learning}}, - url = {http://arxiv.org/abs/1802.04799}, - urldate = {2023-10-26}, - year = 2018, - Bdsk-Url-1 = {http://arxiv.org/abs/1802.04799}} - + title = {{TVM}: {An} {Automated} {End}-to-{End} {Optimizing} {Compiler} for {Deep} {Learning}}, + shorttitle = {{TVM}}, + author = {Chen, Tianqi and Moreau, Thierry and Jiang, Ziheng and Zheng, Lianmin and Yan, Eddie and Cowan, Meghan and Shen, Haichen and Wang, Leyuan and Hu, Yuwei and Ceze, Luis and Guestrin, Carlos and Krishnamurthy, Arvind}, + year = 2018, + month = oct, + booktitle = {13th USENIX Symposium on Operating Systems Design and Implementation (OSDI 18)}, + publisher = {arXiv}, + pages = {578--594}, + url = {http://arxiv.org/abs/1802.04799}, + urldate = {2023-10-26}, + note = {arXiv:1802.04799 [cs]}, + annote = {Comment: Significantly improved version, add automated optimization}, + file = {Chen et al. - 2018 - TVM An Automated End-to-End Optimizing Compiler f.pdf:/Users/alex/Zotero/storage/QR8MHJ38/Chen et al. - 2018 - TVM An Automated End-to-End Optimizing Compiler f.pdf:application/pdf}, + keywords = {Computer Science - Artificial Intelligence, Computer Science - Machine Learning, Computer Science - Programming Languages}, + language = {en}, + bdsk-url-1 = {http://arxiv.org/abs/1802.04799} +} +@article{chen2006gallium, + title = {Gallium, indium, and arsenic pollution of groundwater from a semiconductor manufacturing area of Taiwan.}, + author = {Chen, H-W}, + year = 2006, + journal = {Bulletin of Environmental Contamination \& Toxicology}, + volume = 77, + number = 2 +} @article{chen2016training, - author = {Chen, Tianqi and Xu, Bing and Zhang, Chiyuan and Guestrin, Carlos}, - journal = {arXiv preprint arXiv:1604.06174}, - title = {Training deep nets with sublinear memory cost}, - year = 2016} - + title = {Training deep nets with sublinear memory cost}, + author = {Chen, Tianqi and Xu, Bing and Zhang, Chiyuan and Guestrin, Carlos}, + year = 2016, + journal = {arXiv preprint arXiv:1604.06174} +} @inproceedings{chen2018tvm, - author = {Chen, Tianqi and Moreau, Thierry and Jiang, Ziheng and Zheng, Lianmin and Yan, Eddie and Shen, Haichen and Cowan, Meghan and Wang, Leyuan and Hu, Yuwei and Ceze, Luis and others}, - booktitle = {13th USENIX Symposium on Operating Systems Design and Implementation (OSDI 18)}, - pages = {578--594}, - title = {$\{$TVM$\}$: An automated $\{$End-to-End$\}$ optimizing compiler for deep learning}, - year = 2018} - + title = {$\{$TVM$\}$: An automated $\{$End-to-End$\}$ optimizing compiler for deep learning}, + author = {Chen, Tianqi and Moreau, Thierry and Jiang, Ziheng and Zheng, Lianmin and Yan, Eddie and Shen, Haichen and Cowan, Meghan and Wang, Leyuan and Hu, Yuwei and Ceze, Luis and others}, + year = 2018, + booktitle = {13th USENIX Symposium on Operating Systems Design and Implementation (OSDI 18)}, + pages = {578--594} +} +@article{chen2019looks, + title = {This looks like that: deep learning for interpretable image recognition}, + author = {Chen, Chaofan and Li, Oscar and Tao, Daniel and Barnett, Alina and Rudin, Cynthia and Su, Jonathan K}, + year = 2019, + journal = {Advances in neural information processing systems}, + volume = 32 +} +@article{Chen2023, + title = {A framework for integrating artificial intelligence for clinical care with continuous therapeutic monitoring}, + author = {Chen, Emma and Prakash, Shvetank and Janapa Reddi, Vijay and Kim, David and Rajpurkar, Pranav}, + year = 2023, + month = {Nov}, + day = {06}, + journal = {Nature Biomedical Engineering}, + doi = {10.1038/s41551-023-01115-0}, + issn = {2157-846X}, + url = {https://doi.org/10.1038/s41551-023-01115-0}, + bdsk-url-1 = {https://doi.org/10.1038/s41551-023-01115-0} +} @article{chen2023learning, - author = {Chen, Zhiyong and Xu, Shugong}, - journal = {EURASIP Journal on Audio, Speech, and Music Processing}, - number = 1, - pages = 33, - publisher = {Springer}, - title = {Learning domain-heterogeneous speaker recognition systems with personalized continual federated learning}, - volume = 2023, - year = 2023} - + title = {Learning domain-heterogeneous speaker recognition systems with personalized continual federated learning}, + author = {Chen, Zhiyong and Xu, Shugong}, + year = 2023, + journal = {EURASIP Journal on Audio, Speech, and Music Processing}, + publisher = {Springer}, + volume = 2023, + number = 1, + pages = 33 +} +@article{cheng2017survey, + title = {A survey of model compression and acceleration for deep neural networks}, + author = {Cheng, Yu and Wang, Duo and Zhou, Pan and Zhang, Tao}, + year = 2017, + journal = {arXiv preprint arXiv:1710.09282} +} +@article{chenGalliumIndiumArsenic2006, + title = {Gallium, indium, and arsenic pollution of groundwater from a semiconductor manufacturing area of Taiwan.}, + author = {Chen, H-W}, + year = 2006, + journal = {Bulletin of Environmental Contamination \& Toxicology}, + volume = 77, + number = 2 +} +@article{chi2016prime, + title = {Prime: A novel processing-in-memory architecture for neural network computation in reram-based main memory}, + author = {Chi, Ping and Li, Shuangchen and Xu, Cong and Zhang, Tao and Zhao, Jishen and Liu, Yongpan and Wang, Yu and Xie, Yuan}, + year = 2016, + journal = {ACM SIGARCH Computer Architecture News}, + publisher = {ACM New York, NY, USA}, + volume = 44, + number = 3, + pages = {27--39} +} @misc{chollet2015, - author = {Fran{\c c}ois Chollet}, - commit = {5bcac37}, - howpublished = {\url{https://github.com/fchollet/keras}}, - journal = {GitHub repository}, - publisher = {GitHub}, - title = {keras}, - year = 2015} - + title = {keras}, + author = {Fran{\c c}ois Chollet}, + year = 2015, + journal = {GitHub repository}, + publisher = {GitHub}, + commit = {5bcac37}, + howpublished = {\url{https://github.com/fchollet/keras}} +} @article{chollet2018keras, - author = {Chollet, Fran{\c{c}}ois}, - journal = {March 9th}, - title = {Introduction to keras}, - year = 2018} - - + title = {Introduction to keras}, + author = {Chollet, Fran{\c{c}}ois}, + year = 2018, + journal = {March 9th} +} +@article{christiano2017deep, + title = {Deep reinforcement learning from human preferences}, + author = {Christiano, Paul F and Leike, Jan and Brown, Tom and Martic, Miljan and Legg, Shane and Amodei, Dario}, + year = 2017, + journal = {Advances in neural information processing systems}, + volume = 30 +} @inproceedings{chu2021discovering, + title = {Discovering multi-hardware mobile models via architecture search}, + author = {Chu, Grace and Arikan, Okan and Bender, Gabriel and Wang, Weijun and Brighton, Achille and Kindermans, Pieter-Jan and Liu, Hanxiao and Akin, Berkin and Gupta, Suyog and Howard, Andrew}, + year = 2021, + booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition}, + pages = {3022--3031}, archiveprefix = {arXiv}, - author = {Chu, Grace and Arikan, Okan and Bender, Gabriel and Wang, Weijun and Brighton, Achille and Kindermans, Pieter-Jan and Liu, Hanxiao and Akin, Berkin and Gupta, Suyog and Howard, Andrew}, - booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition}, - eprint = {2008.08178}, - pages = {3022--3031}, - primaryclass = {cs.CV}, - title = {Discovering multi-hardware mobile models via architecture search}, - year = 2021} - + eprint = {2008.08178}, + primaryclass = {cs.CV} +} +@article{chua1971memristor, + title = {Memristor-the missing circuit element}, + author = {Chua, Leon}, + year = 1971, + journal = {IEEE Transactions on circuit theory}, + publisher = {IEEE}, + volume = 18, + number = 5, + pages = {507--519} +} @article{coleman2017dawnbench, - author = {Coleman, Cody and Narayanan, Deepak and Kang, Daniel and Zhao, Tian and Zhang, Jian and Nardi, Luigi and Bailis, Peter and Olukotun, Kunle and R{\'e}, Chris and Zaharia, Matei}, - journal = {Training}, - number = 101, - pages = 102, - title = {Dawnbench: An end-to-end deep learning benchmark and competition}, - volume = 100, - year = 2017} - + title = {Dawnbench: An end-to-end deep learning benchmark and competition}, + author = {Coleman, Cody and Narayanan, Deepak and Kang, Daniel and Zhao, Tian and Zhang, Jian and Nardi, Luigi and Bailis, Peter and Olukotun, Kunle and R{\'e}, Chris and Zaharia, Matei}, + year = 2017, + journal = {Training}, + volume = 100, + number = 101, + pages = 102 +} @inproceedings{coleman2022similarity, - author = {Coleman, Cody and Chou, Edward and Katz-Samuels, Julian and Culatana, Sean and Bailis, Peter and Berg, Alexander C and Nowak, Robert and Sumbaly, Roshan and Zaharia, Matei and Yalniz, I Zeki}, - booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence}, - number = 6, - pages = {6402--6410}, - title = {Similarity search for efficient active learning and search of rare concepts}, - volume = 36, - year = 2022} - + title = {Similarity search for efficient active learning and search of rare concepts}, + author = {Coleman, Cody and Chou, Edward and Katz-Samuels, Julian and Culatana, Sean and Bailis, Peter and Berg, Alexander C and Nowak, Robert and Sumbaly, Roshan and Zaharia, Matei and Yalniz, I Zeki}, + year = 2022, + booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence}, + volume = 36, + number = 6, + pages = {6402--6410} +} +@inproceedings{cooper2011semiconductor, + title = {A semiconductor company's examination of its water footprint approach}, + author = {Cooper, Tom and Fallender, Suzanne and Pafumi, Joyann and Dettling, Jon and Humbert, Sebastien and Lessard, Lindsay}, + year = 2011, + booktitle = {Proceedings of the 2011 IEEE International Symposium on Sustainable Systems and Technology}, + pages = {1--6}, + organization = {IEEE} +} +@article{cope2009pure, + title = {Pure water, semiconductors and the recession}, + author = {Cope, Gord}, + year = 2009, + journal = {Global Water Intelligence}, + volume = 10, + number = 10 +} @misc{cottier_trends_2023, - author = {Cottier, Ben}, - journal = {Epoch AI Report}, - month = jan, - title = {Trends in the {Dollar} {Training} {Cost} of {Machine} {Learning} {Systems}}, - url = {https://epochai.org/blog/trends-in-the-dollar-training-cost-of-machine-learning-systems}, - year = 2023, - Bdsk-Url-1 = {https://epochai.org/blog/trends-in-the-dollar-training-cost-of-machine-learning-systems}} - + title = {Trends in the {Dollar} {Training} {Cost} of {Machine} {Learning} {Systems}}, + author = {Cottier, Ben}, + year = 2023, + month = jan, + journal = {Epoch AI Report}, + url = {https://epochai.org/blog/trends-in-the-dollar-training-cost-of-machine-learning-systems}, + bdsk-url-1 = {https://epochai.org/blog/trends-in-the-dollar-training-cost-of-machine-learning-systems} +} +@book{d2023dataFeminism, + title = {Data feminism}, + author = {D'ignazio, Catherine and Klein, Lauren F}, + year = 2023, + publisher = {MIT press} +} +@article{dahl2023benchmarking, + title = {Benchmarking Neural Network Training Algorithms}, + author = {Dahl, George E and Schneider, Frank and Nado, Zachary and Agarwal, Naman and Sastry, Chandramouli Shama and Hennig, Philipp and Medapati, Sourabh and Eschenhagen, Runa and Kasimbeg, Priya and Suo, Daniel and others}, + year = 2023, + journal = {arXiv preprint arXiv:2306.07179} +} +@article{dally_evolution_2021, + title = {Evolution of the {Graphics} {Processing} {Unit} ({GPU})}, + author = {Dally, William J. and Keckler, Stephen W. and Kirk, David B.}, + year = 2021, + month = nov, + journal = {IEEE Micro}, + volume = 41, + number = 6, + pages = {42--51}, + doi = {10.1109/MM.2021.3113475}, + issn = {1937-4143}, + url = {https://ieeexplore.ieee.org/document/9623445}, + urldate = {2023-11-07}, + note = {Conference Name: IEEE Micro}, + abstract = {Graphics processing units (GPUs) power today's fastest supercomputers, are the dominant platform for deep learning, and provide the intelligence for devices ranging from self-driving cars to robots and smart cameras. They also generate compelling photorealistic images at real-time frame rates. GPUs have evolved by adding features to support new use cases. NVIDIA's GeForce 256, the first GPU, was a dedicated processor for real-time graphics, an application that demands large amounts of floating-point arithmetic for vertex and fragment shading computations and high memory bandwidth. As real-time graphics advanced, GPUs became programmable. The combination of programmability and floating-point performance made GPUs attractive for running scientific applications. Scientists found ways to use early programmable GPUs by casting their calculations as vertex and fragment shaders. GPUs evolved to meet the needs of scientific users by adding hardware for simpler programming, double-precision floating-point arithmetic, and resilience.}, + bdsk-url-1 = {https://ieeexplore.ieee.org/document/9623445}, + bdsk-url-2 = {https://doi.org/10.1109/MM.2021.3113475} +} +@article{data_centers_wheels, + title = {Data Centers on Wheels: Emissions From Computing Onboard Autonomous Vehicles}, + author = {Sudhakar, Soumya and Sze, Vivienne and Karaman, Sertac}, + year = 2023, + journal = {IEEE Micro}, + volume = 43, + number = 1, + pages = {29--39}, + doi = {10.1109/MM.2022.3219803} +} @misc{david_tensorflow_2021, - author = {David, Robert and Duke, Jared and Jain, Advait and Reddi, Vijay Janapa and Jeffries, Nat and Li, Jian and Kreeger, Nick and Nappier, Ian and Natraj, Meghna and Regev, Shlomi and Rhodes, Rocky and Wang, Tiezhen and Warden, Pete}, - file = {David et al. - 2021 - TensorFlow Lite Micro Embedded Machine Learning o.pdf:/Users/alex/Zotero/storage/YCFVNEVH/David et al. - 2021 - TensorFlow Lite Micro Embedded Machine Learning o.pdf:application/pdf}, - keywords = {Computer Science - Artificial Intelligence, Computer Science - Machine Learning}, - language = {en}, - month = mar, - note = {arXiv:2010.08678 [cs]}, - publisher = {arXiv}, - shorttitle = {{TensorFlow} {Lite} {Micro}}, - title = {{TensorFlow} {Lite} {Micro}: {Embedded} {Machine} {Learning} on {TinyML} {Systems}}, - url = {http://arxiv.org/abs/2010.08678}, - urldate = {2023-10-26}, - year = 2021, - Bdsk-Url-1 = {http://arxiv.org/abs/2010.08678}} - + title = {{TensorFlow} {Lite} {Micro}: {Embedded} {Machine} {Learning} on {TinyML} {Systems}}, + shorttitle = {{TensorFlow} {Lite} {Micro}}, + author = {David, Robert and Duke, Jared and Jain, Advait and Reddi, Vijay Janapa and Jeffries, Nat and Li, Jian and Kreeger, Nick and Nappier, Ian and Natraj, Meghna and Regev, Shlomi and Rhodes, Rocky and Wang, Tiezhen and Warden, Pete}, + year = 2021, + month = mar, + publisher = {arXiv}, + url = {http://arxiv.org/abs/2010.08678}, + urldate = {2023-10-26}, + note = {arXiv:2010.08678 [cs]}, + file = {David et al. - 2021 - TensorFlow Lite Micro Embedded Machine Learning o.pdf:/Users/alex/Zotero/storage/YCFVNEVH/David et al. - 2021 - TensorFlow Lite Micro Embedded Machine Learning o.pdf:application/pdf}, + keywords = {Computer Science - Artificial Intelligence, Computer Science - Machine Learning}, + language = {en}, + bdsk-url-1 = {http://arxiv.org/abs/2010.08678} +} @article{david2021tensorflow, - author = {David, Robert and Duke, Jared and Jain, Advait and Janapa Reddi, Vijay and Jeffries, Nat and Li, Jian and Kreeger, Nick and Nappier, Ian and Natraj, Meghna and Wang, Tiezhen and others}, - journal = {Proceedings of Machine Learning and Systems}, - pages = {800--811}, - title = {Tensorflow lite micro: Embedded machine learning for tinyml systems}, - volume = 3, - year = 2021} - + title = {Tensorflow lite micro: Embedded machine learning for tinyml systems}, + author = {David, Robert and Duke, Jared and Jain, Advait and Janapa Reddi, Vijay and Jeffries, Nat and Li, Jian and Kreeger, Nick and Nappier, Ian and Natraj, Meghna and Wang, Tiezhen and others}, + year = 2021, + journal = {Proceedings of Machine Learning and Systems}, + volume = 3, + pages = {800--811} +} +@article{davies2018loihi, + title = {Loihi: A neuromorphic manycore processor with on-chip learning}, + author = {Davies, Mike and Srinivasa, Narayan and Lin, Tsung-Han and Chinya, Gautham and Cao, Yongqiang and Choday, Sri Harsha and Dimou, Georgios and Joshi, Prasad and Imam, Nabil and Jain, Shweta and others}, + year = 2018, + journal = {Ieee Micro}, + publisher = {IEEE}, + volume = 38, + number = 1, + pages = {82--99} +} +@article{davies2021advancing, + title = {Advancing neuromorphic computing with loihi: A survey of results and outlook}, + author = {Davies, Mike and Wild, Andreas and Orchard, Garrick and Sandamirskaya, Yulia and Guerra, Gabriel A Fonseca and Joshi, Prasad and Plank, Philipp and Risbud, Sumedh R}, + year = 2021, + journal = {Proceedings of the IEEE}, + publisher = {IEEE}, + volume = 109, + number = 5, + pages = {911--934} +} +@techreport{daviesEndangeredElements2011, + title = {Endangered elements: Critical thinking}, + author = {Davies, Emma}, + year = 2011, + month = jan, + pages = {50--54}, + url = {https://www.rsc.org/images/Endangered%20Elements%20-%20Critical%20Thinking_tcm18-196054.pdf} +} +@article{dayarathna2015data, + title = {Data center energy consumption modeling: A survey}, + author = {Dayarathna, Miyuru and Wen, Yonggang and Fan, Rui}, + year = 2015, + journal = {IEEE Communications surveys \& tutorials}, + publisher = {IEEE}, + volume = 18, + number = 1, + pages = {732--794} +} +@misc{dean_jeff_numbers_nodate, + title = {Numbers {Everyone} {Should} {Know}}, + author = {Dean. Jeff}, + url = {https://brenocon.com/dean_perf.html}, + urldate = {2023-11-07}, + bdsk-url-1 = {https://brenocon.com/dean_perf.html} +} @article{dean2012large, - author = {Dean, Jeffrey and Corrado, Greg and Monga, Rajat and Chen, Kai and Devin, Matthieu and Mao, Mark and Ranzato, Marc'aurelio and Senior, Andrew and Tucker, Paul and Yang, Ke and others}, - journal = {Advances in neural information processing systems}, - title = {Large scale distributed deep networks}, - volume = 25, - year = 2012} - + title = {Large scale distributed deep networks}, + author = {Dean, Jeffrey and Corrado, Greg and Monga, Rajat and Chen, Kai and Devin, Matthieu and Mao, Mark and Ranzato, Marc'aurelio and Senior, Andrew and Tucker, Paul and Yang, Ke and others}, + year = 2012, + journal = {Advances in neural information processing systems}, + volume = 25 +} @misc{deci, - title = {The Ultimate Guide to Deep Learning Model Quantization and Quantization-Aware Training}, - url = {https://deci.ai/quantization-and-quantization-aware-training/}, - Bdsk-Url-1 = {https://deci.ai/quantization-and-quantization-aware-training/}} - + title = {The Ultimate Guide to Deep Learning Model Quantization and Quantization-Aware Training}, + url = {https://deci.ai/quantization-and-quantization-aware-training/}, + bdsk-url-1 = {https://deci.ai/quantization-and-quantization-aware-training/} +} @misc{deepcompress, - abstract = {Neural networks are both computationally intensive and memory intensive, making them difficult to deploy on embedded systems with limited hardware resources. To address this limitation, we introduce "deep compression", a three stage pipeline: pruning, trained quantization and Huffman coding, that work together to reduce the storage requirement of neural networks by 35x to 49x without affecting their accuracy. Our method first prunes the network by learning only the important connections. Next, we quantize the weights to enforce weight sharing, finally, we apply Huffman coding. After the first two steps we retrain the network to fine tune the remaining connections and the quantized centroids. Pruning, reduces the number of connections by 9x to 13x; Quantization then reduces the number of bits that represent each connection from 32 to 5. On the ImageNet dataset, our method reduced the storage required by AlexNet by 35x, from 240MB to 6.9MB, without loss of accuracy. Our method reduced the size of VGG-16 by 49x from 552MB to 11.3MB, again with no loss of accuracy. This allows fitting the model into on-chip SRAM cache rather than off-chip DRAM memory. Our compression method also facilitates the use of complex neural networks in mobile applications where application size and download bandwidth are constrained. Benchmarked on CPU, GPU and mobile GPU, compressed network has 3x to 4x layerwise speedup and 3x to 7x better energy efficiency.}, - author = {Han and Mao and Dally}, - doi = {10.48550/arXiv.1510.00149}, - title = {Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding}, - url = {https://arxiv.org/abs/1510.00149}, - urldate = {2016-02-15}, - year = 2016, - Bdsk-Url-1 = {https://arxiv.org/abs/1510.00149}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.1510.00149}} - + title = {Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding}, + author = {Han and Mao and Dally}, + year = 2016, + journal = {arXiv preprint arXiv:1510.00149}, + doi = {10.48550/arXiv.1510.00149}, + url = {https://arxiv.org/abs/1510.00149}, + urldate = {2016-02-15}, + abstract = {Neural networks are both computationally intensive and memory intensive, making them difficult to deploy on embedded systems with limited hardware resources. To address this limitation, we introduce "deep compression", a three stage pipeline: pruning, trained quantization and Huffman coding, that work together to reduce the storage requirement of neural networks by 35x to 49x without affecting their accuracy. Our method first prunes the network by learning only the important connections. Next, we quantize the weights to enforce weight sharing, finally, we apply Huffman coding. After the first two steps we retrain the network to fine tune the remaining connections and the quantized centroids. Pruning, reduces the number of connections by 9x to 13x; Quantization then reduces the number of bits that represent each connection from 32 to 5. On the ImageNet dataset, our method reduced the storage required by AlexNet by 35x, from 240MB to 6.9MB, without loss of accuracy. Our method reduced the size of VGG-16 by 49x from 552MB to 11.3MB, again with no loss of accuracy. This allows fitting the model into on-chip SRAM cache rather than off-chip DRAM memory. Our compression method also facilitates the use of complex neural networks in mobile applications where application size and download bandwidth are constrained. Benchmarked on CPU, GPU and mobile GPU, compressed network has 3x to 4x layerwise speedup and 3x to 7x better energy efficiency.}, + bdsk-url-1 = {https://arxiv.org/abs/1510.00149}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.1510.00149} +} +@article{demler_ceva_2020, + title = {{CEVA} {SENSPRO} {FUSES} {AI} {AND} {VECTOR} {DSP}}, + author = {Demler, Mike}, + year = 2020, + language = {en} +} @inproceedings{deng2009imagenet, - added-at = {2018-09-20T15:22:39.000+0200}, - author = {Deng, Jia and Socher, R. and Fei-Fei, Li and Dong, Wei and Li, Kai and Li, Li-Jia}, - biburl = {https://www.bibsonomy.org/bibtex/252793859f5bcbbd3f7f9e5d083160acf/analyst}, - booktitle = {2009 IEEE Conference on Computer Vision and Pattern Recognition(CVPR)}, - description = {ImageNet: A large-scale hierarchical image database}, - doi = {10.1109/CVPR.2009.5206848}, - interhash = {fbfae3e4fe1a81c477ba00efd0d4d977}, - intrahash = {52793859f5bcbbd3f7f9e5d083160acf}, - keywords = {2009 computer-vision cvpr dataset ieee paper}, - month = {06}, - pages = {248--255}, - timestamp = {2018-09-20T15:22:39.000+0200}, - title = {ImageNet: A large-scale hierarchical image database}, - url = {https://ieeexplore.ieee.org/abstract/document/5206848/}, - volume = 00, - year = 2009, - Bdsk-Url-1 = {https://ieeexplore.ieee.org/abstract/document/5206848/}, - Bdsk-Url-2 = {https://doi.org/10.1109/CVPR.2009.5206848}} - + title = {ImageNet: A large-scale hierarchical image database}, + author = {Deng, Jia and Socher, R. and Fei-Fei, Li and Dong, Wei and Li, Kai and Li, Li-Jia}, + year = 2009, + month = {06}, + booktitle = {2009 IEEE Conference on Computer Vision and Pattern Recognition(CVPR)}, + volume = 00, + pages = {248--255}, + doi = {10.1109/CVPR.2009.5206848}, + url = {https://ieeexplore.ieee.org/abstract/document/5206848/}, + added-at = {2018-09-20T15:22:39.000+0200}, + biburl = {https://www.bibsonomy.org/bibtex/252793859f5bcbbd3f7f9e5d083160acf/analyst}, + description = {ImageNet: A large-scale hierarchical image database}, + interhash = {fbfae3e4fe1a81c477ba00efd0d4d977}, + intrahash = {52793859f5bcbbd3f7f9e5d083160acf}, + keywords = {2009 computer-vision cvpr dataset ieee paper}, + timestamp = {2018-09-20T15:22:39.000+0200}, + bdsk-url-1 = {https://ieeexplore.ieee.org/abstract/document/5206848/}, + bdsk-url-2 = {https://doi.org/10.1109/CVPR.2009.5206848} +} @article{desai2016five, - author = {Desai, Tanvi and Ritchie, Felix and Welpton, Richard and others}, - journal = {Economics Working Paper Series}, - pages = 28, - title = {Five Safes: designing data access for research}, - volume = 1601, - year = 2016} - + title = {Five Safes: designing data access for research}, + author = {Desai, Tanvi and Ritchie, Felix and Welpton, Richard and others}, + year = 2016, + journal = {Economics Working Paper Series}, + volume = 1601, + pages = 28 +} @article{desai2020five, - author = {Desai, Tanvi and Ritchie, Felix and Welpton, Richard}, - journal = {URL https://www2. uwe. ac. uk/faculties/bbs/Documents/1601. pdf}, - title = {Five Safes: designing data access for research; 2016}, - year = 2020} - + title = {Five Safes: designing data access for research; 2016}, + author = {Desai, Tanvi and Ritchie, Felix and Welpton, Richard}, + year = 2020, + journal = {URL https://www2. uwe. ac. uk/faculties/bbs/Documents/1601. pdf} +} @article{devlin2018bert, - author = {Devlin, Jacob and Chang, Ming-Wei and Lee, Kenton and Toutanova, Kristina}, - journal = {arXiv preprint arXiv:1810.04805}, - title = {Bert: Pre-training of deep bidirectional transformers for language understanding}, - year = 2018} - + title = {Bert: Pre-training of deep bidirectional transformers for language understanding}, + author = {Devlin, Jacob and Chang, Ming-Wei and Lee, Kenton and Toutanova, Kristina}, + year = 2018, + journal = {arXiv preprint arXiv:1810.04805} +} +@book{dhanjani2015abusing, + title = {Abusing the internet of things: blackouts, freakouts, and stakeouts}, + author = {Dhanjani, Nitesh}, + year = 2015, + publisher = {O'Reilly Media, Inc.}, + date-added = {2023-11-22 17:09:41 -0500}, + date-modified = {2023-11-22 17:10:22 -0500} +} @article{dhar2021survey, - author = {Dhar, Sauptik and Guo, Junyao and Liu, Jiayi and Tripathi, Samarth and Kurup, Unmesh and Shah, Mohak}, - journal = {ACM Transactions on Internet of Things}, - number = 3, - pages = {1--49}, - publisher = {ACM New York, NY, USA}, - title = {A survey of on-device machine learning: An algorithms and learning theory perspective}, - volume = 2, - year = 2021} - + title = {A survey of on-device machine learning: An algorithms and learning theory perspective}, + author = {Dhar, Sauptik and Guo, Junyao and Liu, Jiayi and Tripathi, Samarth and Kurup, Unmesh and Shah, Mohak}, + year = 2021, + journal = {ACM Transactions on Internet of Things}, + publisher = {ACM New York, NY, USA}, + volume = 2, + number = 3, + pages = {1--49} +} +@misc{dodge2022measuring, + title = {Measuring the Carbon Intensity of AI in Cloud Instances}, + author = {Jesse Dodge and Taylor Prewitt and Remi Tachet Des Combes and Erika Odmark and Roy Schwartz and Emma Strubell and Alexandra Sasha Luccioni and Noah A. Smith and Nicole DeCario and Will Buchanan}, + year = 2022, + eprint = {2206.05229}, + archiveprefix = {arXiv}, + primaryclass = {cs.LG} +} @misc{dong2022splitnets, + title = {SplitNets: Designing Neural Architectures for Efficient Distributed Computing on Head-Mounted Systems}, + author = {Xin Dong and Barbara De Salvo and Meng Li and Chiao Liu and Zhongnan Qu and H. T. Kung and Ziyun Li}, + year = 2022, archiveprefix = {arXiv}, - author = {Xin Dong and Barbara De Salvo and Meng Li and Chiao Liu and Zhongnan Qu and H. T. Kung and Ziyun Li}, - eprint = {2204.04705}, - primaryclass = {cs.LG}, - title = {SplitNets: Designing Neural Architectures for Efficient Distributed Computing on Head-Mounted Systems}, - year = 2022} - + eprint = {2204.04705}, + primaryclass = {cs.LG} +} +@article{Dongarra2009-na, + title = {The evolution of high performance computing on system z}, + author = {Dongarra, Jack J}, + year = 2009, + journal = {IBM Journal of Research and Development}, + volume = 53, + pages = {3--4} +} +@article{dropout, + title = {Dropout: A Simple Way to Prevent Neural Networks from Overfitting}, + author = {Nitish Srivastava and Geoffrey Hinton and Alex Krizhevsky and Ilya Sutskever and Ruslan Salakhutdinov}, + year = 2014, + journal = {Journal of Machine Learning Research}, + url = {http://jmlr.org/papers/v15/srivastava14a.html} +} +@article{duarte2022fastml, + title = {FastML Science Benchmarks: Accelerating Real-Time Scientific Edge Machine Learning}, + author = {Duarte, Javier and Tran, Nhan and Hawks, Ben and Herwig, Christian and Muhizi, Jules and Prakash, Shvetank and Reddi, Vijay Janapa}, + year = 2022, + journal = {arXiv preprint arXiv:2207.07958} +} @article{duisterhof2019learning, - author = {Duisterhof, Bardienus P and Krishnan, Srivatsan and Cruz, Jonathan J and Banbury, Colby R and Fu, William and Faust, Aleksandra and de Croon, Guido CHE and Reddi, Vijay Janapa}, - journal = {arXiv preprint arXiv:1909.11236}, - title = {Learning to seek: Autonomous source seeking with deep reinforcement learning onboard a nano drone microcontroller}, - year = 2019} - + title = {Learning to seek: Autonomous source seeking with deep reinforcement learning onboard a nano drone microcontroller}, + author = {Duisterhof, Bardienus P and Krishnan, Srivatsan and Cruz, Jonathan J and Banbury, Colby R and Fu, William and Faust, Aleksandra and de Croon, Guido CHE and Reddi, Vijay Janapa}, + year = 2019, + journal = {arXiv preprint arXiv:1909.11236} +} @inproceedings{duisterhof2021sniffy, - author = {Duisterhof, Bardienus P and Li, Shushuai and Burgu{\'e}s, Javier and Reddi, Vijay Janapa and de Croon, Guido CHE}, - booktitle = {2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)}, - organization = {IEEE}, - pages = {9099--9106}, - title = {Sniffy bug: A fully autonomous swarm of gas-seeking nano quadcopters in cluttered environments}, - year = 2021} - + title = {Sniffy bug: A fully autonomous swarm of gas-seeking nano quadcopters in cluttered environments}, + author = {Duisterhof, Bardienus P and Li, Shushuai and Burgu{\'e}s, Javier and Reddi, Vijay Janapa and de Croon, Guido CHE}, + year = 2021, + booktitle = {2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)}, + pages = {9099--9106}, + organization = {IEEE} +} +@inproceedings{Dwork2006Theory, + title = {Calibrating Noise to Sensitivity in Private Data Analysis}, + author = {Dwork, Cynthia and McSherry, Frank and Nissim, Kobbi and Smith, Adam}, + year = 2006, + booktitle = {Theory of Cryptography}, + publisher = {Springer Berlin Heidelberg}, + address = {Berlin, Heidelberg}, + pages = {265--284}, + date-added = {2023-11-22 18:04:12 -0500}, + date-modified = {2023-11-22 18:05:20 -0500}, + editor = {Halevi, Shai and Rabin, Tal} +} @article{dwork2014algorithmic, - author = {Dwork, Cynthia and Roth, Aaron and others}, - journal = {Foundations and Trends{\textregistered} in Theoretical Computer Science}, - number = {3--4}, - pages = {211--407}, - publisher = {Now Publishers, Inc.}, - title = {The algorithmic foundations of differential privacy}, - volume = 9, - year = 2014} - + title = {The algorithmic foundations of differential privacy}, + author = {Dwork, Cynthia and Roth, Aaron and others}, + year = 2014, + journal = {Foundations and Trends{\textregistered} in Theoretical Computer Science}, + publisher = {Now Publishers, Inc.}, + volume = 9, + number = {3--4}, + pages = {211--407} +} +@article{e_waste, + title = {Disentangling the worldwide web of e-waste and climate change co-benefits}, + author = {Singh, Narendra and Ogunseitan, Oladele A.}, + year = 2022, + month = dec, + journal = {Circular Economy}, + publisher = {Elsevier BV}, + volume = 1, + number = 2, + pages = 100011, + doi = {10.1016/j.cec.2022.100011}, + issn = {2773-1677}, + url = {http://dx.doi.org/10.1016/j.cec.2022.100011} +} +@article{ebrahimi_review_2014, + title = {A review of data center cooling technology, operating conditions and the corresponding low-grade waste heat recovery opportunities}, + author = {Ebrahimi, Khosrow and Jones, Gerard F. and Fleischer, Amy S.}, + year = 2014, + journal = {Renewable and Sustainable Energy Reviews}, + volume = 31, + pages = {622--638}, + doi = {https://doi.org/10.1016/j.rser.2013.12.007}, + issn = {1364-0321}, + url = {https://www.sciencedirect.com/science/article/pii/S1364032113008216}, + abstract = {The depletion of the world's limited reservoirs of fossil fuels, the worldwide impact of global warming and the high cost of energy are among the primary issues driving a renewed interest in the capture and reuse of waste energy. A major source of waste energy is being created by data centers through the increasing demand for cloud based connectivity and performance. In fact, recent figures show that data centers are responsible for more than 2\% of the US total electricity usage. Almost half of this power is used for cooling the electronics, creating a significant stream of waste heat. The difficulty associated with recovering and reusing this stream of waste heat is that the heat is of low quality. In this paper, the most promising methods and technologies for recovering data center low-grade waste heat in an effective and economically reasonable way are identified and discussed. A number of currently available and developmental low-grade waste heat recovery techniques including district/plant/water heating, absorption cooling, direct power generation (piezoelectric and thermoelectric), indirect power generation (steam and organic Rankine cycle), biomass co-location, and desalination/clean water are reviewed along with their operational requirements in order to assess the suitability and effectiveness of each technology for data center applications. Based on a comparison between data centers' operational thermodynamic conditions and the operational requirements of the discussed waste heat recovery techniques, absorption cooling and organic Rankine cycle are found to be among the most promising technologies for data center waste heat reuse.}, + keywords = {Absorption refrigeration, Data center, Organic Rankine cycle, Thermoelectric, Waste energy reuse, Waste heat recovery} +} +@article{el-rayis_reconfigurable_nodate, + title = {Reconfigurable {Architectures} for the {Next} {Generation} of {Mobile} {Device} {Telecommunications} {Systems}}, + author = {El-Rayis, Ahmed Osman}, + language = {en} +} +@article{eldan2023whos, + title = {Who's Harry Potter? Approximate Unlearning in LLMs}, + author = {Ronen Eldan and Mark Russinovich}, + year = 2023, + journal = {arXiv preprint arXiv:2310.02238}, + date-added = {2023-11-22 19:24:35 -0500}, + date-modified = {2023-11-22 19:25:20 -0500} +} @article{electronics12102287, + title = {Reviewing Federated Learning Aggregation Algorithms; Strategies, Contributions, Limitations and Future Perspectives}, + author = {Moshawrab, Mohammad and Adda, Mehdi and Bouzouane, Abdenour and Ibrahim, Hussein and Raad, Ali}, + year = 2023, + journal = {Electronics}, + publisher = {MDPI}, + volume = 12, + number = 10, + pages = 2287, + doi = {10.3390/electronics12102287}, + issn = {2079-9292}, + url = {https://www.mdpi.com/2079-9292/12/10/2287}, article-number = 2287, - author = {Moshawrab, Mohammad and Adda, Mehdi and Bouzouane, Abdenour and Ibrahim, Hussein and Raad, Ali}, - doi = {10.3390/electronics12102287}, - issn = {2079-9292}, - journal = {Electronics}, - number = 10, - title = {Reviewing Federated Learning Aggregation Algorithms; Strategies, Contributions, Limitations and Future Perspectives}, - url = {https://www.mdpi.com/2079-9292/12/10/2287}, - volume = 12, - year = 2023, - Bdsk-Url-1 = {https://www.mdpi.com/2079-9292/12/10/2287}, - Bdsk-Url-2 = {https://doi.org/10.3390/electronics12102287}} - + bdsk-url-1 = {https://www.mdpi.com/2079-9292/12/10/2287}, + bdsk-url-2 = {https://doi.org/10.3390/electronics12102287} +} +@article{EnergyCons_Emission, + title = {Energy consumption and emission mitigation prediction based on data center traffic and PUE for global data centers}, + author = {Liu, Yanan and Wei, Xiaoxia and Xiao, Jinyu and Liu, Zhijie and Xu, Yang and Tian, Yun}, + year = 2020, + month = jun, + journal = {Global Energy Interconnection}, + publisher = {Elsevier BV}, + volume = 3, + number = 3, + pages = {272–282}, + doi = {10.1016/j.gloei.2020.07.008}, + issn = {2096-5117}, + url = {http://dx.doi.org/10.1016/j.gloei.2020.07.008} +} @misc{energyproblem, - author = {ISSCC}, - title = {Computing's energy problem (and what we can do about it)}, - url = {https://ieeexplore.ieee.org/document/6757323}, - urldate = {2014-03-06}, - year = 2014, - Bdsk-Url-1 = {https://ieeexplore.ieee.org/document/6757323}} - + title = {Computing's energy problem (and what we can do about it)}, + author = {ISSCC}, + year = 2014, + url = {https://ieeexplore.ieee.org/document/6757323}, + urldate = {2014-03-06}, + bdsk-url-1 = {https://ieeexplore.ieee.org/document/6757323} +} @article{esteva2017dermatologist, - author = {Esteva, Andre and Kuprel, Brett and Novoa, Roberto A and Ko, Justin and Swetter, Susan M and Blau, Helen M and Thrun, Sebastian}, - journal = {nature}, - number = 7639, - pages = {115--118}, - publisher = {Nature Publishing Group}, - title = {Dermatologist-level classification of skin cancer with deep neural networks}, - volume = 542, - year = 2017} - + title = {Dermatologist-level classification of skin cancer with deep neural networks}, + author = {Esteva, Andre and Kuprel, Brett and Novoa, Roberto A and Ko, Justin and Swetter, Susan M and Blau, Helen M and Thrun, Sebastian}, + year = 2017, + journal = {nature}, + publisher = {Nature Publishing Group}, + volume = 542, + number = 7639, + pages = {115--118} +} +@article{eykholt2018robust, + title = {Robust Physical-World Attacks on Deep Learning Models}, + author = {Kevin Eykholt and Ivan Evtimov and Earlence Fernandes and Bo Li and Amir Rahmati and Chaowei Xiao and Atul Prakash and Tadayoshi Kohno and Dawn Song}, + year = 2018, + journal = {arXiv preprint arXiv:1707.08945}, + date-added = {2023-11-22 16:30:51 -0500}, + date-modified = {2023-11-22 16:31:55 -0500} +} @misc{fahim2021hls4ml, + title = {hls4ml: An Open-Source Codesign Workflow to Empower Scientific Low-Power Machine Learning Devices}, + author = {Farah Fahim and Benjamin Hawks and Christian Herwig and James Hirschauer and Sergo Jindariani and Nhan Tran and Luca P. Carloni and Giuseppe Di Guglielmo and Philip Harris and Jeffrey Krupa and Dylan Rankin and Manuel Blanco Valentin and Josiah Hester and Yingyi Luo and John Mamish and Seda Orgrenci-Memik and Thea Aarrestad and Hamza Javed and Vladimir Loncar and Maurizio Pierini and Adrian Alan Pol and Sioni Summers and Javier Duarte and Scott Hauck and Shih-Chieh Hsu and Jennifer Ngadiuba and Mia Liu and Duc Hoang and Edward Kreinar and Zhenbin Wu}, + year = 2021, archiveprefix = {arXiv}, - author = {Farah Fahim and Benjamin Hawks and Christian Herwig and James Hirschauer and Sergo Jindariani and Nhan Tran and Luca P. Carloni and Giuseppe Di Guglielmo and Philip Harris and Jeffrey Krupa and Dylan Rankin and Manuel Blanco Valentin and Josiah Hester and Yingyi Luo and John Mamish and Seda Orgrenci-Memik and Thea Aarrestad and Hamza Javed and Vladimir Loncar and Maurizio Pierini and Adrian Alan Pol and Sioni Summers and Javier Duarte and Scott Hauck and Shih-Chieh Hsu and Jennifer Ngadiuba and Mia Liu and Duc Hoang and Edward Kreinar and Zhenbin Wu}, - eprint = {2103.05579}, - primaryclass = {cs.LG}, - title = {hls4ml: An Open-Source Codesign Workflow to Empower Scientific Low-Power Machine Learning Devices}, - year = 2021} - + eprint = {2103.05579}, + primaryclass = {cs.LG} +} +@article{farah2005neuroethics, + title = {Neuroethics: the practical and the philosophical}, + author = {Farah, Martha J}, + year = 2005, + journal = {Trends in cognitive sciences}, + publisher = {Elsevier}, + volume = 9, + number = 1, + pages = {34--40} +} +@article{farwell2011stuxnet, + title = {Stuxnet and the future of cyber war}, + author = {Farwell, James P and Rohozinski, Rafal}, + year = 2011, + journal = {Survival}, + volume = 53, + number = 1, + pages = {23--40}, + date-added = {2023-11-22 14:03:31 -0500}, + date-modified = {2023-11-22 14:05:19 -0500} +} +@inproceedings{fowers2018configurable, + title = {A configurable cloud-scale DNN processor for real-time AI}, + author = {Fowers, Jeremy and Ovtcharov, Kalin and Papamichael, Michael and Massengill, Todd and Liu, Ming and Lo, Daniel and Alkalay, Shlomi and Haselman, Michael and Adams, Logan and Ghandi, Mahdi and others}, + year = 2018, + booktitle = {2018 ACM/IEEE 45th Annual International Symposium on Computer Architecture (ISCA)}, + pages = {1--14}, + organization = {IEEE} +} @misc{frankle_lottery_2019, - abstract = {Neural network pruning techniques can reduce the parameter counts of trained networks by over 90\%, decreasing storage requirements and improving computational performance of inference without compromising accuracy. However, contemporary experience is that the sparse architectures produced by pruning are difficult to train from the start, which would similarly improve training performance. We find that a standard pruning technique naturally uncovers subnetworks whose initializations made them capable of training effectively. Based on these results, we articulate the "lottery ticket hypothesis:" dense, randomly-initialized, feed-forward networks contain subnetworks ("winning tickets") that - when trained in isolation - reach test accuracy comparable to the original network in a similar number of iterations. The winning tickets we find have won the initialization lottery: their connections have initial weights that make training particularly effective. We present an algorithm to identify winning tickets and a series of experiments that support the lottery ticket hypothesis and the importance of these fortuitous initializations. We consistently find winning tickets that are less than 10-20\% of the size of several fully-connected and convolutional feed-forward architectures for MNIST and CIFAR10. Above this size, the winning tickets that we find learn faster than the original network and reach higher test accuracy.}, - author = {Frankle, Jonathan and Carbin, Michael}, - doi = {10.48550/arXiv.1803.03635}, - file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/6STHYGW5/Frankle and Carbin - 2019 - The Lottery Ticket Hypothesis Finding Sparse, Tra.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/QGNSCTQB/1803.html:text/html}, - keywords = {Computer Science - Artificial Intelligence, Computer Science - Machine Learning, Computer Science - Neural and Evolutionary Computing}, - month = mar, - note = {arXiv:1803.03635 [cs]}, - publisher = {arXiv}, - shorttitle = {The {Lottery} {Ticket} {Hypothesis}}, - title = {The {Lottery} {Ticket} {Hypothesis}: {Finding} {Sparse}, {Trainable} {Neural} {Networks}}, - url = {http://arxiv.org/abs/1803.03635}, - urldate = {2023-10-20}, - year = 2019, - Bdsk-Url-1 = {http://arxiv.org/abs/1803.03635}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.1803.03635}} - + title = {The {Lottery} {Ticket} {Hypothesis}: {Finding} {Sparse}, {Trainable} {Neural} {Networks}}, + shorttitle = {The {Lottery} {Ticket} {Hypothesis}}, + author = {Frankle, Jonathan and Carbin, Michael}, + year = 2019, + month = mar, + publisher = {arXiv}, + doi = {10.48550/arXiv.1803.03635}, + url = {http://arxiv.org/abs/1803.03635}, + urldate = {2023-10-20}, + note = {arXiv:1803.03635 [cs]}, + abstract = {Neural network pruning techniques can reduce the parameter counts of trained networks by over 90\%, decreasing storage requirements and improving computational performance of inference without compromising accuracy. However, contemporary experience is that the sparse architectures produced by pruning are difficult to train from the start, which would similarly improve training performance. We find that a standard pruning technique naturally uncovers subnetworks whose initializations made them capable of training effectively. Based on these results, we articulate the "lottery ticket hypothesis:" dense, randomly-initialized, feed-forward networks contain subnetworks ("winning tickets") that - when trained in isolation - reach test accuracy comparable to the original network in a similar number of iterations. The winning tickets we find have won the initialization lottery: their connections have initial weights that make training particularly effective. We present an algorithm to identify winning tickets and a series of experiments that support the lottery ticket hypothesis and the importance of these fortuitous initializations. We consistently find winning tickets that are less than 10-20\% of the size of several fully-connected and convolutional feed-forward architectures for MNIST and CIFAR10. Above this size, the winning tickets that we find learn faster than the original network and reach higher test accuracy.}, + file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/6STHYGW5/Frankle and Carbin - 2019 - The Lottery Ticket Hypothesis Finding Sparse, Tra.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/QGNSCTQB/1803.html:text/html}, + keywords = {Computer Science - Artificial Intelligence, Computer Science - Machine Learning, Computer Science - Neural and Evolutionary Computing}, + bdsk-url-1 = {http://arxiv.org/abs/1803.03635}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.1803.03635} +} +@article{friedman1996value, + title = {Value-sensitive design}, + author = {Friedman, Batya}, + year = 1996, + journal = {interactions}, + publisher = {ACM New York, NY, USA}, + volume = 3, + number = 6, + pages = {16--23} +} +@article{furber2016large, + title = {Large-scale neuromorphic computing systems}, + author = {Furber, Steve}, + year = 2016, + journal = {Journal of neural engineering}, + publisher = {IOP Publishing}, + volume = 13, + number = 5, + pages = {051001} +} +@article{gaitathome, + title = {Monitoring gait at home with radio waves in Parkinson's disease: A marker of severity, progression, and medication response}, + author = {Yingcheng Liu and Guo Zhang and Christopher G. Tarolli and Rumen Hristov and Stella Jensen-Roberts and Emma M. Waddell and Taylor L. Myers and Meghan E. Pawlik and Julia M. Soto and Renee M. Wilson and Yuzhe Yang and Timothy Nordahl and Karlo J. Lizarraga and Jamie L. Adams and Ruth B. Schneider and Karl Kieburtz and Terry Ellis and E. Ray Dorsey and Dina Katabi}, + year = 2022, + journal = {Science Translational Medicine}, + volume = 14, + number = 663, + pages = {eadc9669}, + doi = {10.1126/scitranslmed.adc9669}, + url = {https://www.science.org/doi/abs/10.1126/scitranslmed.adc9669}, + eprint = {https://www.science.org/doi/pdf/10.1126/scitranslmed.adc9669}, + bdsk-url-1 = {https://www.science.org/doi/abs/10.1126/scitranslmed.adc9669}, + bdsk-url-2 = {https://doi.org/10.1126/scitranslmed.adc9669} +} +@article{gale2019state, + title = {The state of sparsity in deep neural networks}, + author = {Gale, Trevor and Elsen, Erich and Hooker, Sara}, + year = 2019, + journal = {arXiv preprint arXiv:1902.09574} +} +@inproceedings{gandolfi2001electromagnetic, + title = {Electromagnetic analysis: Concrete results}, + author = {Gandolfi, Karine and Mourtel, Christophe and Olivier, Francis}, + year = 2001, + booktitle = {Cryptographic Hardware and Embedded Systems---CHES 2001: Third International Workshop Paris, France, May 14--16, 2001 Proceedings 3}, + pages = {251--261}, + date-added = {2023-11-22 16:56:42 -0500}, + date-modified = {2023-11-22 16:57:40 -0500}, + organization = {Springer} +} +@inproceedings{gannot1994verilog, + title = {Verilog HDL based FPGA design}, + author = {Gannot, G. and Ligthart, M.}, + year = 1994, + booktitle = {International Verilog HDL Conference}, + volume = {}, + number = {}, + pages = {86--92}, + doi = {10.1109/IVC.1994.323743}, + bdsk-url-1 = {https://doi.org/10.1109/IVC.1994.323743} +} +@article{Gao2020Physical, + title = {Physical unclonable functions}, + author = {Gao, Yansong and Al-Sarawi, Said F. and Abbott, Derek}, + year = 2020, + month = {February}, + journal = {Nature Electronics}, + volume = 3, + number = 2, + pages = {81--91}, + date-added = {2023-11-22 17:52:20 -0500}, + date-modified = {2023-11-22 17:54:56 -0500} +} +@article{gates2009flexible, + title = {Flexible electronics}, + author = {Gates, Byron D}, + year = 2009, + journal = {Science}, + publisher = {American Association for the Advancement of Science}, + volume = 323, + number = 5921, + pages = {1566--1567} +} @article{gaviria2022dollar, - author = {Gaviria Rojas, William and Diamos, Sudnya and Kini, Keertan and Kanter, David and Janapa Reddi, Vijay and Coleman, Cody}, - journal = {Advances in Neural Information Processing Systems}, - pages = {12979--12990}, - title = {The Dollar Street Dataset: Images Representing the Geographic and Socioeconomic Diversity of the World}, - volume = 35, - year = 2022} - + title = {The Dollar Street Dataset: Images Representing the Geographic and Socioeconomic Diversity of the World}, + author = {Gaviria Rojas, William and Diamos, Sudnya and Kini, Keertan and Kanter, David and Janapa Reddi, Vijay and Coleman, Cody}, + year = 2022, + journal = {Advances in Neural Information Processing Systems}, + volume = 35, + pages = {12979--12990} +} @article{Gebru_Morgenstern_Vecchione_Vaughan_Wallach_III_Crawford_2021, - author = {Gebru, Timnit and Morgenstern, Jamie and Vecchione, Briana and Vaughan, Jennifer Wortman and Wallach, Hanna and III, Hal Daum{\'e} and Crawford, Kate}, - doi = {10.1145/3458723}, - journal = {Communications of the ACM}, - number = 12, - pages = {86--92}, - title = {Datasheets for datasets}, - volume = 64, - year = 2021, - Bdsk-Url-1 = {https://doi.org/10.1145/3458723}} - + title = {Datasheets for datasets}, + author = {Gebru, Timnit and Morgenstern, Jamie and Vecchione, Briana and Vaughan, Jennifer Wortman and Wallach, Hanna and III, Hal Daum{\'e} and Crawford, Kate}, + year = 2021, + journal = {Communications of the ACM}, + volume = 64, + number = 12, + pages = {86--92}, + doi = {10.1145/3458723}, + bdsk-url-1 = {https://doi.org/10.1145/3458723} +} +@article{geiger2021causal, + title = {Causal abstractions of neural networks}, + author = {Geiger, Atticus and Lu, Hanson and Icard, Thomas and Potts, Christopher}, + year = 2021, + journal = {Advances in Neural Information Processing Systems}, + volume = 34, + pages = {9574--9586} +} +@article{glucosemonitor, + title = {Non-invasive Monitoring of Three Glucose Ranges Based On ECG By Using DBSCAN-CNN}, + author = {Li, Jingzhen and Tobore, Igbe and Liu, Yuhang and Kandwal, Abhishek and Wang, Lei and Nie, Zedong}, + year = 2021, + journal = {IEEE Journal of Biomedical and Health Informatics}, + volume = 25, + number = 9, + pages = {3340--3350}, + doi = {10.1109/JBHI.2021.3072628}, + bdsk-url-1 = {https://doi.org/10.1109/JBHI.2021.3072628} +} +@inproceedings{gnad2017voltage, + title = {Voltage drop-based fault attacks on FPGAs using valid bitstreams}, + author = {Gnad, Dennis RE and Oboril, Fabian and Tahoori, Mehdi B}, + year = 2017, + booktitle = {2017 27th International Conference on Field Programmable Logic and Applications (FPL)}, + pages = {1--7}, + date-added = {2023-11-22 17:07:13 -0500}, + date-modified = {2023-11-22 17:07:59 -0500}, + organization = {IEEE} +} @article{goodfellow2020generative, - author = {Goodfellow, Ian and Pouget-Abadie, Jean and Mirza, Mehdi and Xu, Bing and Warde-Farley, David and Ozair, Sherjil and Courville, Aaron and Bengio, Yoshua}, - journal = {Communications of the ACM}, - number = 11, - pages = {139--144}, - publisher = {ACM New York, NY, USA}, - title = {Generative adversarial networks}, - volume = 63, - year = 2020} - + title = {Generative adversarial networks}, + author = {Goodfellow, Ian and Pouget-Abadie, Jean and Mirza, Mehdi and Xu, Bing and Warde-Farley, David and Ozair, Sherjil and Courville, Aaron and Bengio, Yoshua}, + year = 2020, + journal = {Communications of the ACM}, + publisher = {ACM New York, NY, USA}, + volume = 63, + number = 11, + pages = {139--144} +} +@article{goodyear2017social, + title = {Social media, apps and wearable technologies: navigating ethical dilemmas and procedures}, + author = {Goodyear, Victoria A}, + year = 2017, + journal = {Qualitative research in sport, exercise and health}, + publisher = {Taylor \& Francis}, + volume = 9, + number = 3, + pages = {285--302} +} @misc{Google, - author = {Google}, - title = {Information quality & content moderation}, - url = {https://blog.google/documents/83/}, - Bdsk-Url-1 = {https://blog.google/documents/83/}} - + title = {Information quality & content moderation}, + author = {Google}, + url = {https://blog.google/documents/83/}, + bdsk-url-1 = {https://blog.google/documents/83/} +} @misc{gordon_morphnet_2018, - abstract = {We present MorphNet, an approach to automate the design of neural network structures. MorphNet iteratively shrinks and expands a network, shrinking via a resource-weighted sparsifying regularizer on activations and expanding via a uniform multiplicative factor on all layers. In contrast to previous approaches, our method is scalable to large networks, adaptable to specific resource constraints (e.g. the number of floating-point operations per inference), and capable of increasing the network's performance. When applied to standard network architectures on a wide variety of datasets, our approach discovers novel structures in each domain, obtaining higher performance while respecting the resource constraint.}, - author = {Gordon, Ariel and Eban, Elad and Nachum, Ofir and Chen, Bo and Wu, Hao and Yang, Tien-Ju and Choi, Edward}, - doi = {10.48550/arXiv.1711.06798}, - file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/GV7N4CZC/Gordon et al. - 2018 - MorphNet Fast & Simple Resource-Constrained Struc.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/K6FUV82F/1711.html:text/html}, - keywords = {Computer Science - Machine Learning, Statistics - Machine Learning}, - month = apr, - note = {arXiv:1711.06798 [cs, stat]}, - publisher = {arXiv}, - shorttitle = {{MorphNet}}, - title = {{MorphNet}: {Fast} \& {Simple} {Resource}-{Constrained} {Structure} {Learning} of {Deep} {Networks}}, - url = {http://arxiv.org/abs/1711.06798}, - urldate = {2023-10-20}, - year = 2018, - Bdsk-Url-1 = {http://arxiv.org/abs/1711.06798}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.1711.06798}} - + title = {{MorphNet}: {Fast} \& {Simple} {Resource}-{Constrained} {Structure} {Learning} of {Deep} {Networks}}, + shorttitle = {{MorphNet}}, + author = {Gordon, Ariel and Eban, Elad and Nachum, Ofir and Chen, Bo and Wu, Hao and Yang, Tien-Ju and Choi, Edward}, + year = 2018, + month = apr, + booktitle = {Proceedings of the IEEE conference on computer vision and pattern recognition}, + publisher = {arXiv}, + pages = {1586--1595}, + doi = {10.48550/arXiv.1711.06798}, + url = {http://arxiv.org/abs/1711.06798}, + urldate = {2023-10-20}, + note = {arXiv:1711.06798 [cs, stat]}, + abstract = {We present MorphNet, an approach to automate the design of neural network structures. MorphNet iteratively shrinks and expands a network, shrinking via a resource-weighted sparsifying regularizer on activations and expanding via a uniform multiplicative factor on all layers. In contrast to previous approaches, our method is scalable to large networks, adaptable to specific resource constraints (e.g. the number of floating-point operations per inference), and capable of increasing the network's performance. When applied to standard network architectures on a wide variety of datasets, our approach discovers novel structures in each domain, obtaining higher performance while respecting the resource constraint.}, + file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/GV7N4CZC/Gordon et al. - 2018 - MorphNet Fast & Simple Resource-Constrained Struc.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/K6FUV82F/1711.html:text/html}, + keywords = {Computer Science - Machine Learning, Statistics - Machine Learning}, + bdsk-url-1 = {http://arxiv.org/abs/1711.06798}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.1711.06798} +} @inproceedings{gordon2018morphnet, - author = {Gordon, Ariel and Eban, Elad and Nachum, Ofir and Chen, Bo and Wu, Hao and Yang, Tien-Ju and Choi, Edward}, - booktitle = {Proceedings of the IEEE conference on computer vision and pattern recognition}, - pages = {1586--1595}, - title = {Morphnet: Fast \& simple resource-constrained structure learning of deep networks}, - year = 2018} - + title = {Morphnet: Fast \& simple resource-constrained structure learning of deep networks}, + author = {Gordon, Ariel and Eban, Elad and Nachum, Ofir and Chen, Bo and Wu, Hao and Yang, Tien-Ju and Choi, Edward}, + year = 2018, + booktitle = {Proceedings of the IEEE conference on computer vision and pattern recognition}, + pages = {1586--1595} +} +@inproceedings{govindavajhala2003using, + title = {Using memory errors to attack a virtual machine}, + author = {Govindavajhala, Sudhakar and Appel, Andrew W}, + year = 2003, + booktitle = {2003 Symposium on Security and Privacy, 2003.}, + pages = {154--156}, + date-added = {2023-11-22 16:46:13 -0500}, + date-modified = {2023-11-22 16:47:03 -0500}, + organization = {IEEE} +} +@article{green_AI, + title = {Green ai}, + author = {Schwartz, Roy and Dodge, Jesse and Smith, Noah A and Etzioni, Oren}, + year = 2020, + journal = {Communications of the ACM}, + publisher = {ACM New York, NY, USA}, + volume = 63, + number = 12, + pages = {54--63} +} +@book{grossmanHighTechTrash2007, + title = {High tech trash: Digital devices, hidden toxics, and human health}, + author = {Grossman, Elizabeth}, + year = 2007, + publisher = {Island press} +} @article{gruslys2016memory, - author = {Gruslys, Audrunas and Munos, R{\'e}mi and Danihelka, Ivo and Lanctot, Marc and Graves, Alex}, - journal = {Advances in neural information processing systems}, - title = {Memory-efficient backpropagation through time}, - volume = 29, - year = 2016} - + title = {Memory-efficient backpropagation through time}, + author = {Gruslys, Audrunas and Munos, R{\'e}mi and Danihelka, Ivo and Lanctot, Marc and Graves, Alex}, + year = 2016, + journal = {Advances in neural information processing systems}, + volume = 29 +} +@article{gu2014towards, + title = {Towards deep neural network architectures robust to adversarial examples}, + author = {Gu, Shixiang and Rigazio, Luca}, + year = 2014, + journal = {arXiv preprint arXiv:1412.5068} +} +@article{gupta2016monotonic, + title = {Monotonic calibrated interpolated look-up tables}, + author = {Gupta, Maya and Cotter, Andrew and Pfeifer, Jan and Voevodski, Konstantin and Canini, Kevin and Mangylov, Alexander and Moczydlowski, Wojciech and Van Esbroeck, Alexander}, + year = 2016, + journal = {The Journal of Machine Learning Research}, + publisher = {JMLR. org}, + volume = 17, + number = 1, + pages = {3790--3836} +} +@inproceedings{gupta2022act, + title = {ACT: Designing Sustainable Computer Systems with an Architectural Carbon Modeling Tool}, + shorttitle = {ACT: designing sustainable computer systems with an architectural carbon modeling tool}, + author = {Gupta, Udit and Elgamal, Mariam and Hills, Gage and Wei, Gu-Yeon and Lee, Hsien-Hsin S. and Brooks, David and Wu, Carole-Jean}, + year = 2022, + month = jun, + booktitle = {Proceedings of the 49th Annual International Symposium on Computer Architecture}, + location = {New York, New York}, + publisher = {Association for Computing Machinery}, + address = {New York, NY, USA}, + series = {ISCA '22}, + pages = {784–799}, + doi = {10.1145/3470496.3527408}, + isbn = 9781450386104, + url = {https://doi.org/10.1145/3470496.3527408}, + urldate = {2023-12-06}, + abstract = {Given the performance and efficiency optimizations realized by the computer systems and architecture community over the last decades, the dominating source of computing's carbon footprint is shifting from operational emissions to embodied emissions. These embodied emissions owe to hardware manufacturing and infrastructure-related activities. Despite the rising embodied emissions, there is a distinct lack of architectural modeling tools to quantify and optimize the end-to-end carbon footprint of computing. This work proposes ACT, an architectural carbon footprint modeling framework, to enable carbon characterization and sustainability-driven early design space exploration. Using ACT we demonstrate optimizing hardware for carbon yields distinct solutions compared to optimizing for performance and efficiency. We construct use cases, based on the three tenets of sustainable design---Reduce, Reuse, Recycle---to highlight future methods that enable strong performance and efficiency scaling in an environmentally sustainable manner.}, + numpages = 16, + keywords = {manufacturing, energy, sustainable computing, mobile, computer architecture}, + language = {en} +} +@article{Gupta2023ChatGPT, + title = {From ChatGPT to ThreatGPT: Impact of Generative AI in Cybersecurity and Privacy}, + author = {Gupta, Maanak and Akiri, Charankumar and Aryal, Kshitiz and Parker, Eli and Praharaj, Lopamudra}, + year = 2023, + journal = {IEEE Access}, + volume = 11, + pages = {80218--80245}, + date-added = {2023-11-22 18:01:41 -0500}, + date-modified = {2023-11-22 18:02:55 -0500} +} +@inproceedings{guptaACTDesigningSustainable2022, + title = {ACT: Designing sustainable computer systems with an architectural carbon modeling tool}, + author = {Gupta, Udit and Elgamal, Mariam and Hills, Gage and Wei, Gu-Yeon and Lee, Hsien-Hsin S and Brooks, David and Wu, Carole-Jean}, + year = 2022, + booktitle = {Proceedings of the 49th Annual International Symposium on Computer Architecture}, + pages = {784--799} +} +@article{gwennap_certus-nx_nodate, + title = {Certus-{NX} {Innovates} {General}-{Purpose} {FPGAs}}, + author = {Gwennap, Linley}, + language = {en} +} +@article{haensch2018next, + title = {The next generation of deep learning hardware: Analog computing}, + author = {Haensch, Wilfried and Gokmen, Tayfun and Puri, Ruchir}, + year = 2018, + journal = {Proceedings of the IEEE}, + publisher = {IEEE}, + volume = 107, + number = 1, + pages = {108--122} +} @article{han2015deep, - author = {Han, Song and Mao, Huizi and Dally, William J}, - journal = {arXiv preprint arXiv:1510.00149}, - title = {Deep compression: Compressing deep neural networks with pruning, trained quantization and huffman coding}, - year = 2015} - + title = {Deep compression: Compressing deep neural networks with pruning, trained quantization and huffman coding}, + author = {Han, Song and Mao, Huizi and Dally, William J}, + year = 2015, + journal = {arXiv preprint arXiv:1510.00149} +} @misc{han2016deep, + title = {Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding}, + author = {Song Han and Huizi Mao and William J. Dally}, + year = 2016, archiveprefix = {arXiv}, - author = {Song Han and Huizi Mao and William J. Dally}, - eprint = {1510.00149}, - primaryclass = {cs.CV}, - title = {Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding}, - year = 2016} - + eprint = {1510.00149}, + primaryclass = {cs.CV} +} +@article{handlin1965science, + title = {Science and technology in popular culture}, + author = {Handlin, Oscar}, + year = 1965, + journal = {Daedalus}, + publisher = {JSTOR}, + pages = {156--170} +} +@article{hardt2016equality, + title = {Equality of opportunity in supervised learning}, + author = {Hardt, Moritz and Price, Eric and Srebro, Nati}, + year = 2016, + journal = {Advances in neural information processing systems}, + volume = 29 +} +@article{hazan2021neuromorphic, + title = {Neuromorphic analog implementation of neural engineering framework-inspired spiking neuron for high-dimensional representation}, + author = {Hazan, Avi and Ezra Tsur, Elishai}, + year = 2021, + journal = {Frontiers in Neuroscience}, + publisher = {Frontiers Media SA}, + volume = 15, + pages = 627221 +} @misc{he_structured_2023, - abstract = {The remarkable performance of deep Convolutional neural networks (CNNs) is generally attributed to their deeper and wider architectures, which can come with significant computational costs. Pruning neural networks has thus gained interest since it effectively lowers storage and computational costs. In contrast to weight pruning, which results in unstructured models, structured pruning provides the benefit of realistic acceleration by producing models that are friendly to hardware implementation. The special requirements of structured pruning have led to the discovery of numerous new challenges and the development of innovative solutions. This article surveys the recent progress towards structured pruning of deep CNNs. We summarize and compare the state-of-the-art structured pruning techniques with respect to filter ranking methods, regularization methods, dynamic execution, neural architecture search, the lottery ticket hypothesis, and the applications of pruning. While discussing structured pruning algorithms, we briefly introduce the unstructured pruning counterpart to emphasize their differences. Furthermore, we provide insights into potential research opportunities in the field of structured pruning. A curated list of neural network pruning papers can be found at https://github.com/he-y/Awesome-Pruning}, - author = {He, Yang and Xiao, Lingao}, - doi = {10.48550/arXiv.2303.00566}, - file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/K5RGQQA9/He and Xiao - 2023 - Structured Pruning for Deep Convolutional Neural N.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/U7PVPU4C/2303.html:text/html}, - keywords = {Computer Science - Computer Vision and Pattern Recognition}, - month = mar, - note = {arXiv:2303.00566 [cs]}, - publisher = {arXiv}, - shorttitle = {Structured {Pruning} for {Deep} {Convolutional} {Neural} {Networks}}, - title = {Structured {Pruning} for {Deep} {Convolutional} {Neural} {Networks}: {A} survey}, - url = {http://arxiv.org/abs/2303.00566}, - urldate = {2023-10-20}, - year = 2023, - Bdsk-Url-1 = {http://arxiv.org/abs/2303.00566}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.2303.00566}} - + title = {Structured {Pruning} for {Deep} {Convolutional} {Neural} {Networks}: {A} survey}, + shorttitle = {Structured {Pruning} for {Deep} {Convolutional} {Neural} {Networks}}, + author = {He, Yang and Xiao, Lingao}, + year = 2023, + month = mar, + publisher = {arXiv}, + doi = {10.48550/arXiv.2303.00566}, + url = {http://arxiv.org/abs/2303.00566}, + urldate = {2023-10-20}, + note = {arXiv:2303.00566 [cs]}, + abstract = {The remarkable performance of deep Convolutional neural networks (CNNs) is generally attributed to their deeper and wider architectures, which can come with significant computational costs. Pruning neural networks has thus gained interest since it effectively lowers storage and computational costs. In contrast to weight pruning, which results in unstructured models, structured pruning provides the benefit of realistic acceleration by producing models that are friendly to hardware implementation. The special requirements of structured pruning have led to the discovery of numerous new challenges and the development of innovative solutions. This article surveys the recent progress towards structured pruning of deep CNNs. We summarize and compare the state-of-the-art structured pruning techniques with respect to filter ranking methods, regularization methods, dynamic execution, neural architecture search, the lottery ticket hypothesis, and the applications of pruning. While discussing structured pruning algorithms, we briefly introduce the unstructured pruning counterpart to emphasize their differences. Furthermore, we provide insights into potential research opportunities in the field of structured pruning. A curated list of neural network pruning papers can be found at https://github.com/he-y/Awesome-Pruning}, + file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/K5RGQQA9/He and Xiao - 2023 - Structured Pruning for Deep Convolutional Neural N.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/U7PVPU4C/2303.html:text/html}, + keywords = {Computer Science - Computer Vision and Pattern Recognition}, + bdsk-url-1 = {http://arxiv.org/abs/2303.00566}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.2303.00566} +} @inproceedings{he2016deep, - author = {He, Kaiming and Zhang, Xiangyu and Ren, Shaoqing and Sun, Jian}, - booktitle = {Proceedings of the IEEE conference on computer vision and pattern recognition}, - pages = {770--778}, - title = {Deep residual learning for image recognition}, - year = 2016} - + title = {Deep residual learning for image recognition}, + author = {He, Kaiming and Zhang, Xiangyu and Ren, Shaoqing and Sun, Jian}, + year = 2016, + booktitle = {Proceedings of the IEEE conference on computer vision and pattern recognition}, + pages = {770--778} +} +@inproceedings{hebert2018multicalibration, + title = {Multicalibration: Calibration for the (computationally-identifiable) masses}, + author = {H{\'e}bert-Johnson, Ursula and Kim, Michael and Reingold, Omer and Rothblum, Guy}, + year = 2018, + booktitle = {International Conference on Machine Learning}, + pages = {1939--1948}, + organization = {PMLR} +} +@article{henderson2020towards, + title = {Towards the systematic reporting of the energy and carbon footprints of machine learning}, + author = {Henderson, Peter and Hu, Jieru and Romoff, Joshua and Brunskill, Emma and Jurafsky, Dan and Pineau, Joelle}, + year = 2020, + journal = {The Journal of Machine Learning Research}, + publisher = {JMLRORG}, + volume = 21, + number = 1, + pages = {10039--10081} +} @inproceedings{hendrycks2021natural, - author = {Hendrycks, Dan and Zhao, Kevin and Basart, Steven and Steinhardt, Jacob and Song, Dawn}, - booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition}, - pages = {15262--15271}, - title = {Natural adversarial examples}, - year = 2021} - + title = {Natural adversarial examples}, + author = {Hendrycks, Dan and Zhao, Kevin and Basart, Steven and Steinhardt, Jacob and Song, Dawn}, + year = 2021, + booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition}, + pages = {15262--15271} +} +@article{Hennessy2019-je, + title = {A new golden age for computer architecture}, + author = {Hennessy, John L and Patterson, David A}, + year = 2019, + month = jan, + journal = {Commun. ACM}, + publisher = {Association for Computing Machinery (ACM)}, + volume = 62, + number = 2, + pages = {48--60}, + copyright = {http://www.acm.org/publications/policies/copyright\_policy\#Background}, + abstract = {Innovations like domain-specific hardware, enhanced security, open instruction sets, and agile chip development will lead the way.}, + language = {en} +} +@article{himmelstein2022examination, + title = {Examination of stigmatizing language in the electronic health record}, + author = {Himmelstein, Gracie and Bates, David and Zhou, Li}, + year = 2022, + journal = {JAMA Network Open}, + publisher = {American Medical Association}, + volume = 5, + number = 1, + pages = {e2144967--e2144967} +} @misc{hinton_distilling_2015, - abstract = {A very simple way to improve the performance of almost any machine learning algorithm is to train many different models on the same data and then to average their predictions. Unfortunately, making predictions using a whole ensemble of models is cumbersome and may be too computationally expensive to allow deployment to a large number of users, especially if the individual models are large neural nets. Caruana and his collaborators have shown that it is possible to compress the knowledge in an ensemble into a single model which is much easier to deploy and we develop this approach further using a different compression technique. We achieve some surprising results on MNIST and we show that we can significantly improve the acoustic model of a heavily used commercial system by distilling the knowledge in an ensemble of models into a single model. We also introduce a new type of ensemble composed of one or more full models and many specialist models which learn to distinguish fine-grained classes that the full models confuse. Unlike a mixture of experts, these specialist models can be trained rapidly and in parallel.}, - author = {Hinton, Geoffrey and Vinyals, Oriol and Dean, Jeff}, - doi = {10.48550/arXiv.1503.02531}, - file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/VREDW45A/Hinton et al. - 2015 - Distilling the Knowledge in a Neural Network.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/8MNJG4RP/1503.html:text/html}, - keywords = {Computer Science - Machine Learning, Computer Science - Neural and Evolutionary Computing, Statistics - Machine Learning}, - month = mar, - note = {arXiv:1503.02531 [cs, stat]}, - publisher = {arXiv}, - title = {Distilling the {Knowledge} in a {Neural} {Network}}, - url = {http://arxiv.org/abs/1503.02531}, - urldate = {2023-10-20}, - year = 2015, - Bdsk-Url-1 = {http://arxiv.org/abs/1503.02531}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.1503.02531}} - + title = {Distilling the {Knowledge} in a {Neural} {Network}}, + author = {Hinton, Geoffrey and Vinyals, Oriol and Dean, Jeff}, + year = 2015, + month = mar, + publisher = {arXiv}, + doi = {10.48550/arXiv.1503.02531}, + url = {http://arxiv.org/abs/1503.02531}, + urldate = {2023-10-20}, + note = {arXiv:1503.02531 [cs, stat]}, + abstract = {A very simple way to improve the performance of almost any machine learning algorithm is to train many different models on the same data and then to average their predictions. Unfortunately, making predictions using a whole ensemble of models is cumbersome and may be too computationally expensive to allow deployment to a large number of users, especially if the individual models are large neural nets. Caruana and his collaborators have shown that it is possible to compress the knowledge in an ensemble into a single model which is much easier to deploy and we develop this approach further using a different compression technique. We achieve some surprising results on MNIST and we show that we can significantly improve the acoustic model of a heavily used commercial system by distilling the knowledge in an ensemble of models into a single model. We also introduce a new type of ensemble composed of one or more full models and many specialist models which learn to distinguish fine-grained classes that the full models confuse. Unlike a mixture of experts, these specialist models can be trained rapidly and in parallel.}, + file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/VREDW45A/Hinton et al. - 2015 - Distilling the Knowledge in a Neural Network.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/8MNJG4RP/1503.html:text/html}, + keywords = {Computer Science - Machine Learning, Computer Science - Neural and Evolutionary Computing, Statistics - Machine Learning}, + bdsk-url-1 = {http://arxiv.org/abs/1503.02531}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.1503.02531} +} @misc{hinton2015distilling, + title = {Distilling the Knowledge in a Neural Network}, + author = {Geoffrey Hinton and Oriol Vinyals and Jeff Dean}, + year = 2015, archiveprefix = {arXiv}, - author = {Geoffrey Hinton and Oriol Vinyals and Jeff Dean}, - eprint = {1503.02531}, - primaryclass = {stat.ML}, - title = {Distilling the Knowledge in a Neural Network}, - year = 2015} - + eprint = {1503.02531}, + primaryclass = {stat.ML} +} @article{Holland_Hosny_Newman_Joseph_Chmielinski_2020, - author = {Holland, Sarah and Hosny, Ahmed and Newman, Sarah and Joseph, Joshua and Chmielinski, Kasia}, - doi = {10.5040/9781509932771.ch-001}, - journal = {Data Protection and Privacy}, - title = {The Dataset Nutrition label}, - year = 2020, - Bdsk-Url-1 = {https://doi.org/10.5040/9781509932771.ch-001}} - + title = {The Dataset Nutrition label}, + author = {Holland, Sarah and Hosny, Ahmed and Newman, Sarah and Joseph, Joshua and Chmielinski, Kasia}, + year = 2020, + journal = {Data Protection and Privacy}, + doi = {10.5040/9781509932771.ch-001}, + bdsk-url-1 = {https://doi.org/10.5040/9781509932771.ch-001} +} @inproceedings{hong2023publishing, - author = {Hong, Sanghyun and Carlini, Nicholas and Kurakin, Alexey}, - booktitle = {2023 IEEE Conference on Secure and Trustworthy Machine Learning (SaTML)}, - organization = {IEEE}, - pages = {271--290}, - title = {Publishing Efficient On-device Models Increases Adversarial Vulnerability}, - year = 2023} - + title = {Publishing Efficient On-device Models Increases Adversarial Vulnerability}, + author = {Hong, Sanghyun and Carlini, Nicholas and Kurakin, Alexey}, + year = 2023, + booktitle = {2023 IEEE Conference on Secure and Trustworthy Machine Learning (SaTML)}, + pages = {271--290}, + organization = {IEEE} +} +@article{hosseini2017deceiving, + title = {Deceiving google's perspective api built for detecting toxic comments}, + author = {Hosseini, Hossein and Kannan, Sreeram and Zhang, Baosen and Poovendran, Radha}, + year = 2017, + journal = {arXiv preprint arXiv:1702.08138}, + date-added = {2023-11-22 16:22:18 -0500}, + date-modified = {2023-11-22 16:23:43 -0500} +} @misc{howard_mobilenets_2017, - abstract = {We present a class of efficient models called MobileNets for mobile and embedded vision applications. MobileNets are based on a streamlined architecture that uses depth-wise separable convolutions to build light weight deep neural networks. We introduce two simple global hyper-parameters that efficiently trade off between latency and accuracy. These hyper-parameters allow the model builder to choose the right sized model for their application based on the constraints of the problem. We present extensive experiments on resource and accuracy tradeoffs and show strong performance compared to other popular models on ImageNet classification. We then demonstrate the effectiveness of MobileNets across a wide range of applications and use cases including object detection, finegrain classification, face attributes and large scale geo-localization.}, - author = {Howard, Andrew G. and Zhu, Menglong and Chen, Bo and Kalenichenko, Dmitry and Wang, Weijun and Weyand, Tobias and Andreetto, Marco and Adam, Hartwig}, - doi = {10.48550/arXiv.1704.04861}, - file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/IJ9P9ID9/Howard et al. - 2017 - MobileNets Efficient Convolutional Neural Network.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/D9TS95GJ/1704.html:text/html}, - keywords = {Computer Science - Computer Vision and Pattern Recognition}, - month = apr, - note = {arXiv:1704.04861 [cs]}, - publisher = {arXiv}, - shorttitle = {{MobileNets}}, - title = {{MobileNets}: {Efficient} {Convolutional} {Neural} {Networks} for {Mobile} {Vision} {Applications}}, - url = {http://arxiv.org/abs/1704.04861}, - urldate = {2023-10-20}, - year = 2017, - Bdsk-Url-1 = {http://arxiv.org/abs/1704.04861}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.1704.04861}} - + title = {{MobileNets}: {Efficient} {Convolutional} {Neural} {Networks} for {Mobile} {Vision} {Applications}}, + shorttitle = {{MobileNets}}, + author = {Howard, Andrew G. and Zhu, Menglong and Chen, Bo and Kalenichenko, Dmitry and Wang, Weijun and Weyand, Tobias and Andreetto, Marco and Adam, Hartwig}, + year = 2017, + month = apr, + publisher = {arXiv}, + doi = {10.48550/arXiv.1704.04861}, + url = {http://arxiv.org/abs/1704.04861}, + urldate = {2023-10-20}, + note = {arXiv:1704.04861 [cs]}, + abstract = {We present a class of efficient models called MobileNets for mobile and embedded vision applications. MobileNets are based on a streamlined architecture that uses depth-wise separable convolutions to build light weight deep neural networks. We introduce two simple global hyper-parameters that efficiently trade off between latency and accuracy. These hyper-parameters allow the model builder to choose the right sized model for their application based on the constraints of the problem. We present extensive experiments on resource and accuracy tradeoffs and show strong performance compared to other popular models on ImageNet classification. We then demonstrate the effectiveness of MobileNets across a wide range of applications and use cases including object detection, finegrain classification, face attributes and large scale geo-localization.}, + file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/IJ9P9ID9/Howard et al. - 2017 - MobileNets Efficient Convolutional Neural Network.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/D9TS95GJ/1704.html:text/html}, + keywords = {Computer Science - Computer Vision and Pattern Recognition}, + bdsk-url-1 = {http://arxiv.org/abs/1704.04861}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.1704.04861} +} @misc{howard2017mobilenets, + title = {MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications}, + author = {Andrew G. Howard and Menglong Zhu and Bo Chen and Dmitry Kalenichenko and Weijun Wang and Tobias Weyand and Marco Andreetto and Hartwig Adam}, + year = 2017, + journal = {arXiv preprint arXiv:1704.04861}, archiveprefix = {arXiv}, - author = {Andrew G. Howard and Menglong Zhu and Bo Chen and Dmitry Kalenichenko and Weijun Wang and Tobias Weyand and Marco Andreetto and Hartwig Adam}, - eprint = {1704.04861}, - journal = {arXiv preprint arXiv:1704.04861}, - primaryclass = {cs.CV}, - title = {MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications}, - year = 2017} - + eprint = {1704.04861}, + primaryclass = {cs.CV} +} +@inproceedings{hsiao2023mavfi, + title = {Mavfi: An end-to-end fault analysis framework with anomaly detection and recovery for micro aerial vehicles}, + author = {Hsiao, Yu-Shun and Wan, Zishen and Jia, Tianyu and Ghosal, Radhika and Mahmoud, Abdulrahman and Raychowdhury, Arijit and Brooks, David and Wei, Gu-Yeon and Reddi, Vijay Janapa}, + year = 2023, + booktitle = {2023 Design, Automation \& Test in Europe Conference \& Exhibition (DATE)}, + pages = {1--6}, + date-added = {2023-11-22 16:54:11 -0500}, + date-modified = {2023-11-22 16:55:12 -0500}, + organization = {IEEE} +} +@article{hsu2016accumulation, + title = {Accumulation of heavy metals and trace elements in fluvial sediments received effluents from traditional and semiconductor industries}, + author = {Hsu, Liang-Ching and Huang, Ching-Yi and Chuang, Yen-Hsun and Chen, Ho-Wen and Chan, Ya-Ting and Teah, Heng Yi and Chen, Tsan-Yao and Chang, Chiung-Fen and Liu, Yu-Ting and Tzou, Yu-Min}, + year = 2016, + journal = {Scientific reports}, + publisher = {Nature Publishing Group UK London}, + volume = 6, + number = 1, + pages = 34250 +} +@article{huang2010pseudo, + title = {Pseudo-CMOS: A design style for low-cost and robust flexible electronics}, + author = {Huang, Tsung-Ching and Fukuda, Kenjiro and Lo, Chun-Ming and Yeh, Yung-Hui and Sekitani, Tsuyoshi and Someya, Takao and Cheng, Kwang-Ting}, + year = 2010, + journal = {IEEE Transactions on Electron Devices}, + publisher = {IEEE}, + volume = 58, + number = 1, + pages = {141--150} +} +@inproceedings{hutter2009contact, + title = {Contact-based fault injections and power analysis on RFID tags}, + author = {Hutter, Michael and Schmidt, Jorn-Marc and Plos, Thomas}, + year = 2009, + booktitle = {2009 European Conference on Circuit Theory and Design}, + pages = {409--412}, + date-added = {2023-11-22 16:43:29 -0500}, + date-modified = {2023-11-22 16:44:30 -0500}, + organization = {IEEE} +} @misc{iandola_squeezenet_2016, - abstract = {Recent research on deep neural networks has focused primarily on improving accuracy. For a given accuracy level, it is typically possible to identify multiple DNN architectures that achieve that accuracy level. With equivalent accuracy, smaller DNN architectures offer at least three advantages: (1) Smaller DNNs require less communication across servers during distributed training. (2) Smaller DNNs require less bandwidth to export a new model from the cloud to an autonomous car. (3) Smaller DNNs are more feasible to deploy on FPGAs and other hardware with limited memory. To provide all of these advantages, we propose a small DNN architecture called SqueezeNet. SqueezeNet achieves AlexNet-level accuracy on ImageNet with 50x fewer parameters. Additionally, with model compression techniques we are able to compress SqueezeNet to less than 0.5MB (510x smaller than AlexNet). The SqueezeNet architecture is available for download here: https://github.com/DeepScale/SqueezeNet}, - author = {Iandola, Forrest N. and Han, Song and Moskewicz, Matthew W. and Ashraf, Khalid and Dally, William J. and Keutzer, Kurt}, - doi = {10.48550/arXiv.1602.07360}, - file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/X3ZX9UTZ/Iandola et al. - 2016 - SqueezeNet AlexNet-level accuracy with 50x fewer .pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/DHI96QVT/1602.html:text/html}, - keywords = {Computer Science - Artificial Intelligence, Computer Science - Computer Vision and Pattern Recognition}, - month = nov, - note = {arXiv:1602.07360 [cs]}, - publisher = {arXiv}, - shorttitle = {{SqueezeNet}}, - title = {{SqueezeNet}: {AlexNet}-level accuracy with 50x fewer parameters and {\textless}0.{5MB} model size}, - url = {http://arxiv.org/abs/1602.07360}, - urldate = {2023-10-20}, - year = 2016, - Bdsk-Url-1 = {http://arxiv.org/abs/1602.07360}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.1602.07360}} - + title = {{SqueezeNet}: {AlexNet}-level accuracy with 50x fewer parameters and {\textless}0.{5MB} model size}, + shorttitle = {{SqueezeNet}}, + author = {Iandola, Forrest N. and Han, Song and Moskewicz, Matthew W. and Ashraf, Khalid and Dally, William J. and Keutzer, Kurt}, + year = 2016, + month = nov, + publisher = {arXiv}, + doi = {10.48550/arXiv.1602.07360}, + url = {http://arxiv.org/abs/1602.07360}, + urldate = {2023-10-20}, + note = {arXiv:1602.07360 [cs]}, + abstract = {Recent research on deep neural networks has focused primarily on improving accuracy. For a given accuracy level, it is typically possible to identify multiple DNN architectures that achieve that accuracy level. With equivalent accuracy, smaller DNN architectures offer at least three advantages: (1) Smaller DNNs require less communication across servers during distributed training. (2) Smaller DNNs require less bandwidth to export a new model from the cloud to an autonomous car. (3) Smaller DNNs are more feasible to deploy on FPGAs and other hardware with limited memory. To provide all of these advantages, we propose a small DNN architecture called SqueezeNet. SqueezeNet achieves AlexNet-level accuracy on ImageNet with 50x fewer parameters. Additionally, with model compression techniques we are able to compress SqueezeNet to less than 0.5MB (510x smaller than AlexNet). The SqueezeNet architecture is available for download here: https://github.com/DeepScale/SqueezeNet}, + file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/X3ZX9UTZ/Iandola et al. - 2016 - SqueezeNet AlexNet-level accuracy with 50x fewer .pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/DHI96QVT/1602.html:text/html}, + keywords = {Computer Science - Artificial Intelligence, Computer Science - Computer Vision and Pattern Recognition}, + bdsk-url-1 = {http://arxiv.org/abs/1602.07360}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.1602.07360} +} @article{iandola2016squeezenet, - author = {Iandola, Forrest N and Han, Song and Moskewicz, Matthew W and Ashraf, Khalid and Dally, William J and Keutzer, Kurt}, - journal = {arXiv preprint arXiv:1602.07360}, - title = {SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and< 0.5 MB model size}, - year = 2016} - + title = {SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and< 0.5 MB model size}, + author = {Iandola, Forrest N and Han, Song and Moskewicz, Matthew W and Ashraf, Khalid and Dally, William J and Keutzer, Kurt}, + year = 2016, + journal = {arXiv preprint arXiv:1602.07360} +} +@article{Ignatov2018-kh, + title = {{AI} Benchmark: Running deep neural networks on Android smartphones}, + author = {Ignatov, Andrey and Timofte, Radu and Chou, William and Wang, Ke and Wu, Max and Hartley, Tim and Van Gool, Luc}, + year = 2018, + booktitle = {Proceedings of the European Conference on Computer Vision (ECCV) Workshops}, + publisher = {arXiv}, + pages = {0--0}, + abstract = {Over the last years, the computational power of mobile devices such as smartphones and tablets has grown dramatically, reaching the level of desktop computers available not long ago. While standard smartphone apps are no longer a problem for them, there is still a group of tasks that can easily challenge even high-end devices, namely running artificial intelligence algorithms. In this paper, we present a study of the current state of deep learning in the Android ecosystem and describe available frameworks, programming models and the limitations of running AI on smartphones. We give an overview of the hardware acceleration resources available on four main mobile chipset platforms: Qualcomm, HiSilicon, MediaTek and Samsung. Additionally, we present the real-world performance results of different mobile SoCs collected with AI Benchmark that are covering all main existing hardware configurations.} +} @inproceedings{ignatov2018ai, - author = {Ignatov, Andrey and Timofte, Radu and Chou, William and Wang, Ke and Wu, Max and Hartley, Tim and Van Gool, Luc}, - booktitle = {Proceedings of the European Conference on Computer Vision (ECCV) Workshops}, - pages = {0--0}, - title = {Ai benchmark: Running deep neural networks on android smartphones}, - year = 2018} - + title = {Ai benchmark: Running deep neural networks on android smartphones}, + author = {Ignatov, Andrey and Timofte, Radu and Chou, William and Wang, Ke and Wu, Max and Hartley, Tim and Van Gool, Luc}, + year = 2018, + booktitle = {Proceedings of the European Conference on Computer Vision (ECCV) Workshops}, + pages = {0--0} +} @inproceedings{ijcai2021p592, - author = {Benmeziane, Hadjer and El Maghraoui, Kaoutar and Ouarnoughi, Hamza and Niar, Smail and Wistuba, Martin and Wang, Naigang}, - booktitle = {Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence, {IJCAI-21}}, - doi = {10.24963/ijcai.2021/592}, - editor = {Zhi-Hua Zhou}, - month = 8, - note = {Survey Track}, - pages = {4322--4329}, - publisher = {International Joint Conferences on Artificial Intelligence Organization}, - title = {Hardware-Aware Neural Architecture Search: Survey and Taxonomy}, - url = {https://doi.org/10.24963/ijcai.2021/592}, - year = 2021, - Bdsk-Url-1 = {https://doi.org/10.24963/ijcai.2021/592}} - + title = {Hardware-Aware Neural Architecture Search: Survey and Taxonomy}, + author = {Benmeziane, Hadjer and El Maghraoui, Kaoutar and Ouarnoughi, Hamza and Niar, Smail and Wistuba, Martin and Wang, Naigang}, + year = 2021, + month = 8, + booktitle = {Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence, {IJCAI-21}}, + publisher = {International Joint Conferences on Artificial Intelligence Organization}, + pages = {4322--4329}, + doi = {10.24963/ijcai.2021/592}, + url = {https://doi.org/10.24963/ijcai.2021/592}, + note = {Survey Track}, + editor = {Zhi-Hua Zhou}, + bdsk-url-1 = {https://doi.org/10.24963/ijcai.2021/592} +} +@inproceedings{imani2016resistive, + title = {Resistive configurable associative memory for approximate computing}, + author = {Imani, Mohsen and Rahimi, Abbas and Rosing, Tajana S}, + year = 2016, + booktitle = {2016 Design, Automation \& Test in Europe Conference \& Exhibition (DATE)}, + pages = {1327--1332}, + organization = {IEEE} +} @misc{intquantfordeepinf, - author = {Wu and Judd, Zhang and Isaev, Micikevicius}, - doi = {10.48550/arXiv.2004.09602}, - title = {Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation)}, - url = {https://arxiv.org/abs/2004.09602}, - urldate = {2020-04-20}, - year = 2020, - Bdsk-Url-1 = {https://arxiv.org/abs/2004.09602}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.2004.09602}} - + title = {Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation)}, + author = {Wu and Judd, Zhang and Isaev, Micikevicius}, + year = 2020, + doi = {10.48550/arXiv.2004.09602}, + url = {https://arxiv.org/abs/2004.09602}, + urldate = {2020-04-20}, + bdsk-url-1 = {https://arxiv.org/abs/2004.09602}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.2004.09602} +} +@article{irimia-vladu_green_2014, + title = {“{Green}” electronics: biodegradable and biocompatible materials and devices for sustainable future}, + author = {Irimia-Vladu, Mihai}, + year = 2014, + journal = {Chem. Soc. Rev.}, + volume = 43, + number = 2, + pages = {588--610}, + doi = {10.1039/C3CS60235D}, + url = {http://dx.doi.org/10.1039/C3CS60235D}, + note = {Publisher: The Royal Society of Chemistry}, + abstract = {“Green” electronics represents not only a novel scientific term but also an emerging area of research aimed at identifying compounds of natural origin and establishing economically efficient routes for the production of synthetic materials that have applicability in environmentally safe (biodegradable) and/or biocompatible devices. The ultimate goal of this research is to create paths for the production of human- and environmentally friendly electronics in general and the integration of such electronic circuits with living tissue in particular. Researching into the emerging class of “green” electronics may help fulfill not only the original promise of organic electronics that is to deliver low-cost and energy efficient materials and devices but also achieve unimaginable functionalities for electronics, for example benign integration into life and environment. This Review will highlight recent research advancements in this emerging group of materials and their integration in unconventional organic electronic devices.} +} +@inproceedings{jacob2018quantization, + title = {Quantization and training of neural networks for efficient integer-arithmetic-only inference}, + author = {Jacob, Benoit and Kligys, Skirmantas and Chen, Bo and Zhu, Menglong and Tang, Matthew and Howard, Andrew and Adam, Hartwig and Kalenichenko, Dmitry}, + year = 2018, + booktitle = {Proceedings of the IEEE conference on computer vision and pattern recognition}, + pages = {2704--2713} +} +@article{janapa2023edge, + title = {Edge Impulse: An MLOps Platform for Tiny Machine Learning}, + author = {Janapa Reddi, Vijay and Elium, Alexander and Hymel, Shawn and Tischler, David and Situnayake, Daniel and Ward, Carl and Moreau, Louis and Plunkett, Jenny and Kelcey, Matthew and Baaijens, Mathijs and others}, + year = 2023, + journal = {Proceedings of Machine Learning and Systems}, + volume = 5 +} +@book{jha2014rare, + title = {Rare earth materials: properties and applications}, + author = {Jha, Asu Ram}, + year = 2014, + publisher = {CRC Press} +} +@book{jhaRareEarthMaterials2014, + title = {Rare earth materials: properties and applications}, + author = {Jha, Asu Ram}, + year = 2014, + publisher = {CRC Press} +} +@misc{jia_dissecting_2018, + title = {Dissecting the {NVIDIA} {Volta} {GPU} {Architecture} via {Microbenchmarking}}, + author = {Jia, Zhe and Maggioni, Marco and Staiger, Benjamin and Scarpazza, Daniele P.}, + year = 2018, + month = apr, + publisher = {arXiv}, + url = {http://arxiv.org/abs/1804.06826}, + urldate = {2023-11-07}, + note = {arXiv:1804.06826 [cs]}, + abstract = {Every year, novel NVIDIA GPU designs are introduced. This rapid architectural and technological progression, coupled with a reluctance by manufacturers to disclose low-level details, makes it difficult for even the most proficient GPU software designers to remain up-to-date with the technological advances at a microarchitectural level. To address this dearth of public, microarchitectural-level information on the novel NVIDIA GPUs, independent researchers have resorted to microbenchmarks-based dissection and discovery. This has led to a prolific line of publications that shed light on instruction encoding, and memory hierarchy's geometry and features at each level. Namely, research that describes the performance and behavior of the Kepler, Maxwell and Pascal architectures. In this technical report, we continue this line of research by presenting the microarchitectural details of the NVIDIA Volta architecture, discovered through microbenchmarks and instruction set disassembly. Additionally, we compare quantitatively our Volta findings against its predecessors, Kepler, Maxwell and Pascal.}, + keywords = {Computer Science - Distributed, Parallel, and Cluster Computing, Computer Science - Performance}, + bdsk-url-1 = {http://arxiv.org/abs/1804.06826} +} @inproceedings{jia2014caffe, - author = {Jia, Yangqing and Shelhamer, Evan and Donahue, Jeff and Karayev, Sergey and Long, Jonathan and Girshick, Ross and Guadarrama, Sergio and Darrell, Trevor}, - booktitle = {Proceedings of the 22nd ACM international conference on Multimedia}, - pages = {675--678}, - title = {Caffe: Convolutional architecture for fast feature embedding}, - year = 2014} - + title = {Caffe: Convolutional architecture for fast feature embedding}, + author = {Jia, Yangqing and Shelhamer, Evan and Donahue, Jeff and Karayev, Sergey and Long, Jonathan and Girshick, Ross and Guadarrama, Sergio and Darrell, Trevor}, + year = 2014, + booktitle = {Proceedings of the 22nd ACM international conference on Multimedia}, + pages = {675--678} +} +@article{jia2019beyond, + title = {Beyond Data and Model Parallelism for Deep Neural Networks.}, + author = {Jia, Zhihao and Zaharia, Matei and Aiken, Alex}, + year = 2019, + journal = {Proceedings of Machine Learning and Systems}, + volume = 1, + pages = {1--13} +} @article{jia2023life, - author = {Jia, Zhenge and Li, Dawei and Xu, Xiaowei and Li, Na and Hong, Feng and Ping, Lichuan and Shi, Yiyu}, - journal = {Nature Machine Intelligence}, - number = 5, - pages = {554--555}, - publisher = {Nature Publishing Group UK London}, - title = {Life-threatening ventricular arrhythmia detection challenge in implantable cardioverter--defibrillators}, - volume = 5, - year = 2023} - + title = {Life-threatening ventricular arrhythmia detection challenge in implantable cardioverter--defibrillators}, + author = {Jia, Zhenge and Li, Dawei and Xu, Xiaowei and Li, Na and Hong, Feng and Ping, Lichuan and Shi, Yiyu}, + year = 2023, + journal = {Nature Machine Intelligence}, + publisher = {Nature Publishing Group UK London}, + volume = 5, + number = 5, + pages = {554--555} +} @misc{jiang2019accuracy, + title = {Accuracy vs. Efficiency: Achieving Both through FPGA-Implementation Aware Neural Architecture Search}, + author = {Weiwen Jiang and Xinyi Zhang and Edwin H. -M. Sha and Lei Yang and Qingfeng Zhuge and Yiyu Shi and Jingtong Hu}, + year = 2019, archiveprefix = {arXiv}, - author = {Weiwen Jiang and Xinyi Zhang and Edwin H. -M. Sha and Lei Yang and Qingfeng Zhuge and Yiyu Shi and Jingtong Hu}, - eprint = {1901.11211}, - primaryclass = {cs.DC}, - title = {Accuracy vs. Efficiency: Achieving Both through FPGA-Implementation Aware Neural Architecture Search}, - year = 2019} - + eprint = {1901.11211}, + primaryclass = {cs.DC} +} @article{Johnson-Roberson_Barto_Mehta_Sridhar_Rosaen_Vasudevan_2017, - author = {Johnson-Roberson, Matthew and Barto, Charles and Mehta, Rounak and Sridhar, Sharath Nittur and Rosaen, Karl and Vasudevan, Ram}, - doi = {10.1109/icra.2017.7989092}, - journal = {2017 IEEE International Conference on Robotics and Automation (ICRA)}, - title = {Driving in the matrix: Can virtual worlds replace human-generated annotations for real world tasks?}, - year = 2017, - Bdsk-Url-1 = {https://doi.org/10.1109/icra.2017.7989092}} - + title = {Driving in the matrix: Can virtual worlds replace human-generated annotations for real world tasks?}, + author = {Johnson-Roberson, Matthew and Barto, Charles and Mehta, Rounak and Sridhar, Sharath Nittur and Rosaen, Karl and Vasudevan, Ram}, + year = 2017, + journal = {2017 IEEE International Conference on Robotics and Automation (ICRA)}, + doi = {10.1109/icra.2017.7989092}, + bdsk-url-1 = {https://doi.org/10.1109/icra.2017.7989092} +} @article{jordan_machine_2015, - author = {Jordan, M. I. and Mitchell, T. M.}, - doi = {10.1126/science.aaa8415}, - file = {Jordan and Mitchell - 2015 - Machine learning Trends, perspectives, and prospe.pdf:/Users/alex/Zotero/storage/RGU3CQ4Q/Jordan and Mitchell - 2015 - Machine learning Trends, perspectives, and prospe.pdf:application/pdf}, - issn = {0036-8075, 1095-9203}, - journal = {Science}, - language = {en}, - month = jul, - number = 6245, - pages = {255--260}, - shorttitle = {Machine learning}, - title = {Machine learning: {Trends}, perspectives, and prospects}, - url = {https://www.science.org/doi/10.1126/science.aaa8415}, - urldate = {2023-10-25}, - volume = 349, - year = 2015, - Bdsk-Url-1 = {https://www.science.org/doi/10.1126/science.aaa8415}, - Bdsk-Url-2 = {https://doi.org/10.1126/science.aaa8415}} - + title = {Machine learning: {Trends}, perspectives, and prospects}, + shorttitle = {Machine learning}, + author = {Jordan, M. I. and Mitchell, T. M.}, + year = 2015, + month = jul, + journal = {Science}, + volume = 349, + number = 6245, + pages = {255--260}, + doi = {10.1126/science.aaa8415}, + issn = {0036-8075, 1095-9203}, + url = {https://www.science.org/doi/10.1126/science.aaa8415}, + urldate = {2023-10-25}, + file = {Jordan and Mitchell - 2015 - Machine learning Trends, perspectives, and prospe.pdf:/Users/alex/Zotero/storage/RGU3CQ4Q/Jordan and Mitchell - 2015 - Machine learning Trends, perspectives, and prospe.pdf:application/pdf}, + language = {en}, + bdsk-url-1 = {https://www.science.org/doi/10.1126/science.aaa8415}, + bdsk-url-2 = {https://doi.org/10.1126/science.aaa8415} +} @inproceedings{jouppi2017datacenter, - author = {Jouppi, Norman P and Young, Cliff and Patil, Nishant and Patterson, David and Agrawal, Gaurav and Bajwa, Raminder and Bates, Sarah and Bhatia, Suresh and Boden, Nan and Borchers, Al and others}, - booktitle = {Proceedings of the 44th annual international symposium on computer architecture}, - pages = {1--12}, - title = {In-datacenter performance analysis of a tensor processing unit}, - year = 2017} - + title = {In-datacenter performance analysis of a tensor processing unit}, + author = {Jouppi, Norman P and Young, Cliff and Patil, Nishant and Patterson, David and Agrawal, Gaurav and Bajwa, Raminder and Bates, Sarah and Bhatia, Suresh and Boden, Nan and Borchers, Al and others}, + year = 2017, + booktitle = {Proceedings of the 44th annual international symposium on computer architecture}, + location = {Toronto, ON, Canada}, + publisher = {Association for Computing Machinery}, + address = {New York, NY, USA}, + series = {ISCA '17}, + pages = {1--12}, + doi = {10.1145/3079856.3080246}, + isbn = 9781450348928, + url = {https://doi.org/10.1145/3079856.3080246}, + abstract = {Many architects believe that major improvements in cost-energy-performance must now come from domain-specific hardware. This paper evaluates a custom ASIC---called a Tensor Processing Unit (TPU) --- deployed in datacenters since 2015 that accelerates the inference phase of neural networks (NN). The heart of the TPU is a 65,536 8-bit MAC matrix multiply unit that offers a peak throughput of 92 TeraOps/second (TOPS) and a large (28 MiB) software-managed on-chip memory. The TPU's deterministic execution model is a better match to the 99th-percentile response-time requirement of our NN applications than are the time-varying optimizations of CPUs and GPUs that help average throughput more than guaranteed latency. The lack of such features helps explain why, despite having myriad MACs and a big memory, the TPU is relatively small and low power. We compare the TPU to a server-class Intel Haswell CPU and an Nvidia K80 GPU, which are contemporaries deployed in the same datacenters. Our workload, written in the high-level TensorFlow framework, uses production NN applications (MLPs, CNNs, and LSTMs) that represent 95\% of our datacenters' NN inference demand. Despite low utilization for some applications, the TPU is on average about 15X -- 30X faster than its contemporary GPU or CPU, with TOPS/Watt about 30X -- 80X higher. Moreover, using the CPU's GDDR5 memory in the TPU would triple achieved TOPS and raise TOPS/Watt to nearly 70X the GPU and 200X the CPU.}, + keywords = {accelerator, neural network, MLP, TPU, CNN, deep learning, domain-specific architecture, GPU, TensorFlow, DNN, RNN, LSTM}, + numpages = 12, + bdsk-url-1 = {https://doi.org/10.1145/3079856.3080246} +} +@inproceedings{Jouppi2023TPUv4, + title = {TPU v4: An Optically Reconfigurable Supercomputer for Machine Learning with Hardware Support for Embeddings}, + author = {Jouppi, Norm and Kurian, George and Li, Sheng and Ma, Peter and Nagarajan, Rahul and Nai, Lifeng and Patil, Nishant and Subramanian, Suvinay and Swing, Andy and Towles, Brian and Young, Clifford and Zhou, Xiang and Zhou, Zongwei and Patterson, David A}, + year = 2023, + booktitle = {Proceedings of the 50th Annual International Symposium on Computer Architecture}, + location = {Orlando, FL, USA}, + publisher = {Association for Computing Machinery}, + address = {New York, NY, USA}, + series = {ISCA '23}, + doi = {10.1145/3579371.3589350}, + isbn = 9798400700958, + url = {https://doi.org/10.1145/3579371.3589350}, + abstract = {In response to innovations in machine learning (ML) models, production workloads changed radically and rapidly. TPU v4 is the fifth Google domain specific architecture (DSA) and its third supercomputer for such ML models. Optical circuit switches (OCSes) dynamically reconfigure its interconnect topology to improve scale, availability, utilization, modularity, deployment, security, power, and performance; users can pick a twisted 3D torus topology if desired. Much cheaper, lower power, and faster than Infiniband, OCSes and underlying optical components are <5\% of system cost and <3\% of system power. Each TPU v4 includes SparseCores, dataflow processors that accelerate models that rely on embeddings by 5x--7x yet use only 5\% of die area and power. Deployed since 2020, TPU v4 outperforms TPU v3 by 2.1x and improves performance/Watt by 2.7x. The TPU v4 supercomputer is 4x larger at 4096 chips and thus nearly 10x faster overall, which along with OCS flexibility and availability allows a large language model to train at an average of ~60\% of peak FLOPS/second. For similar sized systems, it is ~4.3x--4.5x faster than the Graphcore IPU Bow and is 1.2x--1.7x faster and uses 1.3x--1.9x less power than the Nvidia A100. TPU v4s inside the energy-optimized warehouse scale computers of Google Cloud use ~2--6x less energy and produce ~20x less CO2e than contemporary DSAs in typical on-premise data centers.}, + articleno = 82, + keywords = {warehouse scale computer, embeddings, supercomputer, domain specific architecture, reconfigurable, TPU, large language model, power usage effectiveness, CO2 equivalent emissions, energy, optical interconnect, IPU, machine learning, GPU, carbon emissions}, + numpages = 14, + bdsk-url-1 = {https://doi.org/10.1145/3579371.3589350} +} +@book{joye2012fault, + title = {Fault Analysis in Cryptography}, + author = {Joye, Marc and Tunstall, Michael}, + year = 2012, + publisher = {Springer Publishing Company, Incorporated}, + date-added = {2023-11-22 16:35:24 -0500}, + date-modified = {2023-11-22 16:36:20 -0500} +} +@misc{kaiming, + title = {Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification}, + author = {Kaiming He and Xiangyu Zhang and Shaoqing Ren and Jian Sun}, + year = 2015, + booktitle = {Proceedings of the IEEE international conference on computer vision}, + eprint = {1502.01852}, + archiveprefix = {arXiv} +} @article{kairouz2015secure, - author = {Kairouz, Peter and Oh, Sewoong and Viswanath, Pramod}, - journal = {Advances in neural information processing systems}, - title = {Secure multi-party differential privacy}, - volume = 28, - year = 2015} - + title = {Secure multi-party differential privacy}, + author = {Kairouz, Peter and Oh, Sewoong and Viswanath, Pramod}, + year = 2015, + journal = {Advances in neural information processing systems}, + volume = 28 +} @article{karargyris2023federated, - author = {Karargyris, Alexandros and Umeton, Renato and Sheller, Micah J and Aristizabal, Alejandro and George, Johnu and Wuest, Anna and Pati, Sarthak and Kassem, Hasan and Zenk, Maximilian and Baid, Ujjwal and others}, - journal = {Nature Machine Intelligence}, - number = 7, - pages = {799--810}, - publisher = {Nature Publishing Group UK London}, - title = {Federated benchmarking of medical artificial intelligence with MedPerf}, - volume = 5, - year = 2023} - + title = {Federated benchmarking of medical artificial intelligence with MedPerf}, + author = {Karargyris, Alexandros and Umeton, Renato and Sheller, Micah J and Aristizabal, Alejandro and George, Johnu and Wuest, Anna and Pati, Sarthak and Kassem, Hasan and Zenk, Maximilian and Baid, Ujjwal and others}, + year = 2023, + journal = {Nature Machine Intelligence}, + publisher = {Nature Publishing Group UK London}, + volume = 5, + number = 7, + pages = {799--810} +} +@inproceedings{kaur2020interpreting, + title = {Interpreting interpretability: understanding data scientists' use of interpretability tools for machine learning}, + author = {Kaur, Harmanpreet and Nori, Harsha and Jenkins, Samuel and Caruana, Rich and Wallach, Hanna and Wortman Vaughan, Jennifer}, + year = 2020, + booktitle = {Proceedings of the 2020 CHI conference on human factors in computing systems}, + pages = {1--14} +} +@article{khan2021knowledgeadaptation, + title = {Knowledge-Adaptation Priors}, + author = {Mohammad Emtiyaz Khan and Siddharth Swaroop}, + year = 2021, + journal = {arXiv preprint arXiv:2106.08769}, + date-added = {2023-11-22 19:22:50 -0500}, + date-modified = {2023-11-22 19:23:40 -0500} +} @article{kiela2021dynabench, - author = {Kiela, Douwe and Bartolo, Max and Nie, Yixin and Kaushik, Divyansh and Geiger, Atticus and Wu, Zhengxuan and Vidgen, Bertie and Prasad, Grusha and Singh, Amanpreet and Ringshia, Pratik and others}, - journal = {arXiv preprint arXiv:2104.14337}, - title = {Dynabench: Rethinking benchmarking in NLP}, - year = 2021} - + title = {Dynabench: Rethinking benchmarking in NLP}, + author = {Kiela, Douwe and Bartolo, Max and Nie, Yixin and Kaushik, Divyansh and Geiger, Atticus and Wu, Zhengxuan and Vidgen, Bertie and Prasad, Grusha and Singh, Amanpreet and Ringshia, Pratik and others}, + year = 2021, + journal = {arXiv preprint arXiv:2104.14337} +} +@article{kim2018chemical, + title = {Chemical use in the semiconductor manufacturing industry}, + author = {Kim, Sunju and Yoon, Chungsik and Ham, Seunghon and Park, Jihoon and Kwon, Ohun and Park, Donguk and Choi, Sangjun and Kim, Seungwon and Ha, Kwonchul and Kim, Won}, + year = 2018, + journal = {International journal of occupational and environmental health}, + publisher = {Taylor \& Francis}, + volume = 24, + number = {3-4}, + pages = {109--118} +} +@inproceedings{kim2018interpretability, + title = {Interpretability beyond feature attribution: Quantitative testing with concept activation vectors (tcav)}, + author = {Kim, Been and Wattenberg, Martin and Gilmer, Justin and Cai, Carrie and Wexler, James and Viegas, Fernanda and others}, + year = 2018, + booktitle = {International conference on machine learning}, + pages = {2668--2677}, + organization = {PMLR} +} +@article{kimChemicalUseSemiconductor2018, + title = {Chemical use in the semiconductor manufacturing industry}, + author = {Kim, Sunju and Yoon, Chungsik and Ham, Seunghon and Park, Jihoon and Kwon, Ohun and Park, Donguk and Choi, Sangjun and Kim, Seungwon and Ha, Kwonchul and Kim, Won}, + year = 2018, + journal = {International journal of occupational and environmental health}, + publisher = {Taylor \& Francis}, + volume = 24, + number = {3-4}, + pages = {109--118} +} +@inproceedings{kocher1996timing, + title = {Timing attacks on implementations of Diffie-Hellman, RSA, DSS, and other systems}, + author = {Kocher, Paul C}, + year = 1996, + booktitle = {Advances in Cryptology—CRYPTO’96: 16th Annual International Cryptology Conference Santa Barbara, California, USA August 18--22, 1996 Proceedings 16}, + pages = {104--113}, + organization = {Springer} +} +@inproceedings{kocher1999differential, + title = {Differential power analysis}, + author = {Kocher, Paul and Jaffe, Joshua and Jun, Benjamin}, + year = 1999, + booktitle = {Advances in Cryptology---CRYPTO'99: 19th Annual International Cryptology Conference Santa Barbara, California, USA, August 15--19, 1999 Proceedings 19}, + pages = {388--397}, + date-added = {2023-11-22 16:55:28 -0500}, + date-modified = {2023-11-22 16:56:18 -0500}, + organization = {Springer} +} +@article{Kocher2011Intro, + title = {Introduction to differential power analysis}, + author = {Kocher, Paul and Jaffe, Joshua and Jun, Benjamin and Rohatgi, Pankaj}, + year = 2011, + month = {April}, + journal = {Journal of Cryptographic Engineering}, + volume = 1, + number = 1, + pages = {5--27}, + date-added = {2023-11-22 16:58:42 -0500}, + date-modified = {2023-11-22 17:00:36 -0500} +} +@inproceedings{Kocher2018spectre, + title = {Spectre Attacks: Exploiting Speculative Execution}, + author = {Paul Kocher and Jann Horn and Anders Fogh and and Daniel Genkin and Daniel Gruss and Werner Haas and Mike Hamburg and Moritz Lipp and Stefan Mangard and Thomas Prescher and Michael Schwarz and Yuval Yarom}, + year = 2019, + booktitle = {40th IEEE Symposium on Security and Privacy (S\&P'19)}, + date-added = {2023-11-22 16:33:35 -0500}, + date-modified = {2023-11-22 16:34:01 -0500} +} +@inproceedings{koh2020concept, + title = {Concept bottleneck models}, + author = {Koh, Pang Wei and Nguyen, Thao and Tang, Yew Siang and Mussmann, Stephen and Pierson, Emma and Kim, Been and Liang, Percy}, + year = 2020, + booktitle = {International conference on machine learning}, + pages = {5338--5348}, + organization = {PMLR} +} @inproceedings{koh2021wilds, - author = {Koh, Pang Wei and Sagawa, Shiori and Marklund, Henrik and Xie, Sang Michael and Zhang, Marvin and Balsubramani, Akshay and Hu, Weihua and Yasunaga, Michihiro and Phillips, Richard Lanas and Gao, Irena and others}, - booktitle = {International Conference on Machine Learning}, - organization = {PMLR}, - pages = {5637--5664}, - title = {Wilds: A benchmark of in-the-wild distribution shifts}, - year = 2021} - + title = {Wilds: A benchmark of in-the-wild distribution shifts}, + author = {Koh, Pang Wei and Sagawa, Shiori and Marklund, Henrik and Xie, Sang Michael and Zhang, Marvin and Balsubramani, Akshay and Hu, Weihua and Yasunaga, Michihiro and Phillips, Richard Lanas and Gao, Irena and others}, + year = 2021, + booktitle = {International Conference on Machine Learning}, + pages = {5637--5664}, + organization = {PMLR} +} @article{kolda_tensor_2009, - abstract = {This survey provides an overview of higher-order tensor decompositions, their applications, and available software. A tensor is a multidimensional or N -way array. Decompositions of higher-order tensors (i.e., N -way arrays with N ≥ 3) have applications in psychometrics, chemometrics, signal processing, numerical linear algebra, computer vision, numerical analysis, data mining, neuroscience, graph analysis, and elsewhere. Two particular tensor decompositions can be considered to be higher-order extensions of the matrix singular value decomposition: CANDECOMP/PARAFAC (CP) decomposes a tensor as a sum of rank-one tensors, and the Tucker decomposition is a higher-order form of principal component analysis. There are many other tensor decompositions, including INDSCAL, PARAFAC2, CANDELINC, DEDICOM, and PARATUCK2 as well as nonnegative variants of all of the above. The N-way Toolbox, Tensor Toolbox, and Multilinear Engine are examples of software packages for working with tensors.}, - author = {Kolda, Tamara G. and Bader, Brett W.}, - doi = {10.1137/07070111X}, - file = {Kolda and Bader - 2009 - Tensor Decompositions and Applications.pdf:/Users/jeffreyma/Zotero/storage/Q7ZG2267/Kolda and Bader - 2009 - Tensor Decompositions and Applications.pdf:application/pdf}, - issn = {0036-1445, 1095-7200}, - journal = {SIAM Review}, - language = {en}, - month = aug, - number = 3, - pages = {455--500}, - title = {Tensor {Decompositions} and {Applications}}, - url = {http://epubs.siam.org/doi/10.1137/07070111X}, - urldate = {2023-10-20}, - volume = 51, - year = 2009, - Bdsk-Url-1 = {http://epubs.siam.org/doi/10.1137/07070111X}, - Bdsk-Url-2 = {https://doi.org/10.1137/07070111X}} - + title = {Tensor {Decompositions} and {Applications}}, + author = {Kolda, Tamara G. and Bader, Brett W.}, + year = 2009, + month = aug, + journal = {SIAM Review}, + volume = 51, + number = 3, + pages = {455--500}, + doi = {10.1137/07070111X}, + issn = {0036-1445, 1095-7200}, + url = {http://epubs.siam.org/doi/10.1137/07070111X}, + urldate = {2023-10-20}, + abstract = {This survey provides an overview of higher-order tensor decompositions, their applications, and available software. A tensor is a multidimensional or N -way array. Decompositions of higher-order tensors (i.e., N -way arrays with N ≥ 3) have applications in psychometrics, chemometrics, signal processing, numerical linear algebra, computer vision, numerical analysis, data mining, neuroscience, graph analysis, and elsewhere. Two particular tensor decompositions can be considered to be higher-order extensions of the matrix singular value decomposition: CANDECOMP/PARAFAC (CP) decomposes a tensor as a sum of rank-one tensors, and the Tucker decomposition is a higher-order form of principal component analysis. There are many other tensor decompositions, including INDSCAL, PARAFAC2, CANDELINC, DEDICOM, and PARATUCK2 as well as nonnegative variants of all of the above. The N-way Toolbox, Tensor Toolbox, and Multilinear Engine are examples of software packages for working with tensors.}, + file = {Kolda and Bader - 2009 - Tensor Decompositions and Applications.pdf:/Users/jeffreyma/Zotero/storage/Q7ZG2267/Kolda and Bader - 2009 - Tensor Decompositions and Applications.pdf:application/pdf}, + language = {en}, + bdsk-url-1 = {http://epubs.siam.org/doi/10.1137/07070111X}, + bdsk-url-2 = {https://doi.org/10.1137/07070111X} +} @article{koshti2011cumulative, - author = {Koshti, VV}, - journal = {International journal of physics and mathematical sciences}, - number = 1, - pages = {28--32}, - title = {Cumulative sum control chart}, - volume = 1, - year = 2011} - + title = {Cumulative sum control chart}, + author = {Koshti, VV}, + year = 2011, + journal = {International journal of physics and mathematical sciences}, + volume = 1, + number = 1, + pages = {28--32} +} @misc{krishna2023raman, + title = {RAMAN: A Re-configurable and Sparse tinyML Accelerator for Inference on Edge}, + author = {Adithya Krishna and Srikanth Rohit Nudurupati and Chandana D G and Pritesh Dwivedi and Andr{\'e} van Schaik and Mahesh Mehendale and Chetan Singh Thakur}, + year = 2023, archiveprefix = {arXiv}, - author = {Adithya Krishna and Srikanth Rohit Nudurupati and Chandana D G and Pritesh Dwivedi and Andr{\'e} van Schaik and Mahesh Mehendale and Chetan Singh Thakur}, - eprint = {2306.06493}, - primaryclass = {cs.NE}, - title = {RAMAN: A Re-configurable and Sparse tinyML Accelerator for Inference on Edge}, - year = 2023} - + eprint = {2306.06493}, + primaryclass = {cs.NE} +} @article{krishnamoorthi2018quantizing, - author = {Krishnamoorthi, Raghuraman}, - journal = {arXiv preprint arXiv:1806.08342}, - title = {Quantizing deep convolutional networks for efficient inference: A whitepaper}, - year = 2018} - + title = {Quantizing deep convolutional networks for efficient inference: A whitepaper}, + author = {Krishnamoorthi, Raghuraman}, + year = 2018, + month = jun, + journal = {arXiv preprint arXiv:1806.08342}, + publisher = {arXiv}, + doi = {10.48550/arXiv.1806.08342}, + url = {https://arxiv.org/abs/1806.08342}, + urldate = {2018-06-21}, + bdsk-url-1 = {https://arxiv.org/abs/1806.08342}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.1806.08342} +} @article{Krishnan_Rajpurkar_Topol_2022, - author = {Krishnan, Rayan and Rajpurkar, Pranav and Topol, Eric J.}, - doi = {10.1038/s41551-022-00914-1}, - journal = {Nature Biomedical Engineering}, - number = 12, - pages = {1346--1352}, - title = {Self-supervised learning in medicine and Healthcare}, - volume = 6, - year = 2022, - Bdsk-Url-1 = {https://doi.org/10.1038/s41551-022-00914-1}} - + title = {Self-supervised learning in medicine and Healthcare}, + author = {Krishnan, Rayan and Rajpurkar, Pranav and Topol, Eric J.}, + year = 2022, + journal = {Nature Biomedical Engineering}, + volume = 6, + number = 12, + pages = {1346--1352}, + doi = {10.1038/s41551-022-00914-1}, + bdsk-url-1 = {https://doi.org/10.1038/s41551-022-00914-1} +} +@inproceedings{krishnan2023archgym, + title = {ArchGym: An Open-Source Gymnasium for Machine Learning Assisted Architecture Design}, + author = {Krishnan, Srivatsan and Yazdanbakhsh, Amir and Prakash, Shvetank and Jabbour, Jason and Uchendu, Ikechukwu and Ghosh, Susobhan and Boroujerdian, Behzad and Richins, Daniel and Tripathy, Devashree and Faust, Aleksandra and Janapa Reddi, Vijay}, + year = 2023, + booktitle = {Proceedings of the 50th Annual International Symposium on Computer Architecture}, + pages = {1--16} +} @article{krizhevsky2012imagenet, - author = {Krizhevsky, Alex and Sutskever, Ilya and Hinton, Geoffrey E}, - journal = {Advances in neural information processing systems}, - title = {Imagenet classification with deep convolutional neural networks}, - volume = 25, - year = 2012} - + title = {Imagenet classification with deep convolutional neural networks}, + author = {Krizhevsky, Alex and Sutskever, Ilya and Hinton, Geoffrey E}, + year = 2012, + journal = {Advances in neural information processing systems}, + volume = 25 +} @inproceedings{kung1979systolic, - author = {Kung, Hsiang Tsung and Leiserson, Charles E}, - booktitle = {Sparse Matrix Proceedings 1978}, - organization = {Society for industrial and applied mathematics Philadelphia, PA, USA}, - pages = {256--282}, - title = {Systolic arrays (for VLSI)}, - volume = 1, - year = 1979} - + title = {Systolic arrays (for VLSI)}, + author = {Kung, Hsiang Tsung and Leiserson, Charles E}, + year = 1979, + booktitle = {Sparse Matrix Proceedings 1978}, + volume = 1, + pages = {256--282}, + organization = {Society for industrial and applied mathematics Philadelphia, PA, USA} +} @misc{kung2018packing, + title = {Packing Sparse Convolutional Neural Networks for Efficient Systolic Array Implementations: Column Combining Under Joint Optimization}, + author = {H. T. Kung and Bradley McDanel and Sai Qian Zhang}, + year = 2018, archiveprefix = {arXiv}, - author = {H. T. Kung and Bradley McDanel and Sai Qian Zhang}, - eprint = {1811.04770}, - primaryclass = {cs.LG}, - title = {Packing Sparse Convolutional Neural Networks for Efficient Systolic Array Implementations: Column Combining Under Joint Optimization}, - year = 2018} - + eprint = {1811.04770}, + primaryclass = {cs.LG} +} @incollection{kurkova_survey_2018, - address = {Cham}, - author = {Tan, Chuanqi and Sun, Fuchun and Kong, Tao and Zhang, Wenchang and Yang, Chao and Liu, Chunfang}, - booktitle = {Artificial {Neural} {Networks} and {Machine} {Learning} -- {ICANN} 2018}, - doi = {10.1007/978-3-030-01424-7_27}, - editor = {K{\r u}rkov{\'a}, V{\v e}ra and Manolopoulos, Yannis and Hammer, Barbara and Iliadis, Lazaros and Maglogiannis, Ilias}, - file = {Tan et al. - 2018 - A Survey on Deep Transfer Learning.pdf:/Users/alex/Zotero/storage/5NZ36SGB/Tan et al. - 2018 - A Survey on Deep Transfer Learning.pdf:application/pdf}, - isbn = {978-3-030-01423-0 978-3-030-01424-7}, - language = {en}, - note = {Series Title: Lecture Notes in Computer Science}, - pages = {270--279}, - publisher = {Springer International Publishing}, - title = {A {Survey} on {Deep} {Transfer} {Learning}}, - url = {http://link.springer.com/10.1007/978-3-030-01424-7_27}, - urldate = {2023-10-26}, - volume = 11141, - year = 2018, - Bdsk-Url-1 = {http://link.springer.com/10.1007/978-3-030-01424-7_27}, - Bdsk-Url-2 = {https://doi.org/10.1007/978-3-030-01424-7_27}} - + title = {A {Survey} on {Deep} {Transfer} {Learning}}, + author = {Tan, Chuanqi and Sun, Fuchun and Kong, Tao and Zhang, Wenchang and Yang, Chao and Liu, Chunfang}, + year = 2018, + booktitle = {Artificial {Neural} {Networks} and {Machine} {Learning} -- {ICANN} 2018}, + publisher = {Springer International Publishing}, + address = {Cham}, + volume = 11141, + pages = {270--279}, + doi = {10.1007/978-3-030-01424-7_27}, + isbn = {978-3-030-01423-0 978-3-030-01424-7}, + url = {http://link.springer.com/10.1007/978-3-030-01424-7_27}, + urldate = {2023-10-26}, + note = {Series Title: Lecture Notes in Computer Science}, + editor = {K{\r u}rkov{\'a}, V{\v e}ra and Manolopoulos, Yannis and Hammer, Barbara and Iliadis, Lazaros and Maglogiannis, Ilias}, + file = {Tan et al. - 2018 - A Survey on Deep Transfer Learning.pdf:/Users/alex/Zotero/storage/5NZ36SGB/Tan et al. - 2018 - A Survey on Deep Transfer Learning.pdf:application/pdf}, + language = {en}, + bdsk-url-1 = {http://link.springer.com/10.1007/978-3-030-01424-7_27}, + bdsk-url-2 = {https://doi.org/10.1007/978-3-030-01424-7_27} +} +@inproceedings{kurth2023fourcastnet, + title = {Fourcastnet: Accelerating global high-resolution weather forecasting using adaptive fourier neural operators}, + author = {Kurth, Thorsten and Subramanian, Shashank and Harrington, Peter and Pathak, Jaideep and Mardani, Morteza and Hall, David and Miele, Andrea and Kashinath, Karthik and Anandkumar, Anima}, + year = 2023, + booktitle = {Proceedings of the Platform for Advanced Scientific Computing Conference}, + pages = {1--11} +} @misc{kuzmin2022fp8, + title = {FP8 Quantization: The Power of the Exponent}, + author = {Andrey Kuzmin and Mart Van Baalen and Yuwei Ren and Markus Nagel and Jorn Peters and Tijmen Blankevoort}, + year = 2022, archiveprefix = {arXiv}, - author = {Andrey Kuzmin and Mart Van Baalen and Yuwei Ren and Markus Nagel and Jorn Peters and Tijmen Blankevoort}, - eprint = {2208.09225}, - primaryclass = {cs.LG}, - title = {FP8 Quantization: The Power of the Exponent}, - year = 2022} - + eprint = {2208.09225}, + primaryclass = {cs.LG} +} @misc{kwon_tinytrain_2023, - author = {Kwon, Young D. and Li, Rui and Venieris, Stylianos I. and Chauhan, Jagmohan and Lane, Nicholas D. and Mascolo, Cecilia}, - file = {Kwon et al. - 2023 - TinyTrain Deep Neural Network Training at the Ext.pdf:/Users/alex/Zotero/storage/L2ST472U/Kwon et al. - 2023 - TinyTrain Deep Neural Network Training at the Ext.pdf:application/pdf}, - keywords = {Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning}, - language = {en}, - month = jul, - note = {arXiv:2307.09988 [cs]}, - publisher = {arXiv}, - shorttitle = {{TinyTrain}}, - title = {{TinyTrain}: {Deep} {Neural} {Network} {Training} at the {Extreme} {Edge}}, - url = {http://arxiv.org/abs/2307.09988}, - urldate = {2023-10-26}, - year = 2023, - Bdsk-Url-1 = {http://arxiv.org/abs/2307.09988}} - + title = {{TinyTrain}: {Deep} {Neural} {Network} {Training} at the {Extreme} {Edge}}, + shorttitle = {{TinyTrain}}, + author = {Kwon, Young D. and Li, Rui and Venieris, Stylianos I. and Chauhan, Jagmohan and Lane, Nicholas D. and Mascolo, Cecilia}, + year = 2023, + month = jul, + journal = {arXiv preprint arXiv:2307.09988}, + publisher = {arXiv}, + url = {http://arxiv.org/abs/2307.09988}, + urldate = {2023-10-26}, + note = {arXiv:2307.09988 [cs]}, + file = {Kwon et al. - 2023 - TinyTrain Deep Neural Network Training at the Ext.pdf:/Users/alex/Zotero/storage/L2ST472U/Kwon et al. - 2023 - TinyTrain Deep Neural Network Training at the Ext.pdf:application/pdf}, + keywords = {Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning}, + language = {en}, + bdsk-url-1 = {http://arxiv.org/abs/2307.09988} +} +@article{kwon2022flexible, + title = {Flexible sensors and machine learning for heart monitoring}, + author = {Kwon, Sun Hwa and Dong, Lin}, + year = 2022, + journal = {Nano Energy}, + publisher = {Elsevier}, + pages = 107632 +} @article{kwon2023tinytrain, - author = {Kwon, Young D and Li, Rui and Venieris, Stylianos I and Chauhan, Jagmohan and Lane, Nicholas D and Mascolo, Cecilia}, - journal = {arXiv preprint arXiv:2307.09988}, - title = {TinyTrain: Deep Neural Network Training at the Extreme Edge}, - year = 2023} - + title = {TinyTrain: Deep Neural Network Training at the Extreme Edge}, + author = {Kwon, Young D and Li, Rui and Venieris, Stylianos I and Chauhan, Jagmohan and Lane, Nicholas D and Mascolo, Cecilia}, + year = 2023, + journal = {arXiv preprint arXiv:2307.09988} +} @misc{Labelbox, - journal = {Labelbox}, - url = {https://labelbox.com/}, - Bdsk-Url-1 = {https://labelbox.com/}} - + journal = {Labelbox}, + url = {https://labelbox.com/}, + bdsk-url-1 = {https://labelbox.com/} +} @article{lai2018cmsis, - author = {Lai, Liangzhen and Suda, Naveen and Chandra, Vikas}, - journal = {arXiv preprint arXiv:1801.06601}, - title = {Cmsis-nn: Efficient neural network kernels for arm cortex-m cpus}, - year = 2018} - + title = {Cmsis-nn: Efficient neural network kernels for arm cortex-m cpus}, + author = {Lai, Liangzhen and Suda, Naveen and Chandra, Vikas}, + year = 2018, + journal = {arXiv preprint arXiv:1801.06601} +} @misc{lai2018cmsisnn, + title = {CMSIS-NN: Efficient Neural Network Kernels for Arm Cortex-M CPUs}, + author = {Liangzhen Lai and Naveen Suda and Vikas Chandra}, + year = 2018, archiveprefix = {arXiv}, - author = {Liangzhen Lai and Naveen Suda and Vikas Chandra}, - eprint = {1801.06601}, - primaryclass = {cs.NE}, - title = {CMSIS-NN: Efficient Neural Network Kernels for Arm Cortex-M CPUs}, - year = 2018} - + eprint = {1801.06601}, + primaryclass = {cs.NE} +} +@inproceedings{lakkaraju2020fool, + title = {" How do I fool you?" Manipulating User Trust via Misleading Black Box Explanations}, + author = {Lakkaraju, Himabindu and Bastani, Osbert}, + year = 2020, + booktitle = {Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society}, + pages = {79--85} +} +@article{lam2023learning, + title = {Learning skillful medium-range global weather forecasting}, + author = {Lam, Remi and Sanchez-Gonzalez, Alvaro and Willson, Matthew and Wirnsberger, Peter and Fortunato, Meire and Alet, Ferran and Ravuri, Suman and Ewalds, Timo and Eaton-Rosen, Zach and Hu, Weihua and others}, + year = 2023, + journal = {Science}, + publisher = {American Association for the Advancement of Science}, + pages = {eadi2336} +} +@article{lannelongueGreenAlgorithmsQuantifying2020, + shorttitle = {Green {Algorithms}}, + author = {Lannelongue, Loïc and Grealey, Jason and Inouye, Michael}, + year = 2020, + doi = {10.48550/ARXIV.2007.07610}, + url = {https://arxiv.org/abs/2007.07610}, + urldate = {2023-12-06}, + copyright = {Creative Commons Attribution 4.0 International}, + note = {Publisher: arXiv Version Number: 5} +} +@article{lannelongueGreenAlgorithmsQuantifying2021, + shorttitle = {Green {Algorithms}}, + author = {Lannelongue, Loïc and Grealey, Jason and Inouye, Michael}, + year = 2021, + month = jun, + journal = {Advanced Science}, + volume = 8, + number = 12, + pages = 2100707, + doi = {10.1002/advs.202100707}, + issn = {2198-3844, 2198-3844}, + url = {https://onlinelibrary.wiley.com/doi/10.1002/advs.202100707}, + urldate = {2023-12-01}, + language = {en} +} +@article{lecocq2022mitigation, + title = {Mitigation and development pathways in the near-to mid-term (Chapter 4)}, + author = {Lecocq, F and Winkler, H and Daka, JP and Fu, S and Gerber, GS and Kartha, S and Krey, V and Lofgren, H and Masui, T and Mathur, R and others}, + year = 2022, + publisher = {Cambridge University Press} +} @inproceedings{lecun_optimal_1989, - abstract = {We have used information-theoretic ideas to derive a class of prac(cid:173) tical and nearly optimal schemes for adapting the size of a neural network. By removing unimportant weights from a network, sev(cid:173) eral improvements can be expected: better generalization, fewer training examples required, and improved speed of learning and/or classification. The basic idea is to use second-derivative informa(cid:173) tion to make a tradeoff between network complexity and training set error. Experiments confirm the usefulness of the methods on a real-world application.}, - author = {LeCun, Yann and Denker, John and Solla, Sara}, - booktitle = {Advances in {Neural} {Information} {Processing} {Systems}}, - file = {Full Text PDF:/Users/jeffreyma/Zotero/storage/BYHQQSST/LeCun et al. - 1989 - Optimal Brain Damage.pdf:application/pdf}, - publisher = {Morgan-Kaufmann}, - title = {Optimal {Brain} {Damage}}, - url = {https://proceedings.neurips.cc/paper/1989/hash/6c9882bbac1c7093bd25041881277658-Abstract.html}, - urldate = {2023-10-20}, - volume = 2, - year = 1989, - Bdsk-Url-1 = {https://proceedings.neurips.cc/paper/1989/hash/6c9882bbac1c7093bd25041881277658-Abstract.html}} - + title = {Optimal {Brain} {Damage}}, + author = {LeCun, Yann and Denker, John and Solla, Sara}, + year = 1989, + journal = {Advances in neural information processing systems}, + booktitle = {Advances in {Neural} {Information} {Processing} {Systems}}, + publisher = {Morgan-Kaufmann}, + volume = 2, + url = {https://proceedings.neurips.cc/paper/1989/hash/6c9882bbac1c7093bd25041881277658-Abstract.html}, + urldate = {2023-10-20}, + abstract = {We have used information-theoretic ideas to derive a class of prac(cid:173) tical and nearly optimal schemes for adapting the size of a neural network. By removing unimportant weights from a network, sev(cid:173) eral improvements can be expected: better generalization, fewer training examples required, and improved speed of learning and/or classification. The basic idea is to use second-derivative informa(cid:173) tion to make a tradeoff between network complexity and training set error. Experiments confirm the usefulness of the methods on a real-world application.}, + file = {Full Text PDF:/Users/jeffreyma/Zotero/storage/BYHQQSST/LeCun et al. - 1989 - Optimal Brain Damage.pdf:application/pdf}, + bdsk-url-1 = {https://proceedings.neurips.cc/paper/1989/hash/6c9882bbac1c7093bd25041881277658-Abstract.html} +} @article{lecun1989optimal, - author = {LeCun, Yann and Denker, John and Solla, Sara}, - journal = {Advances in neural information processing systems}, - title = {Optimal brain damage}, - volume = 2, - year = 1989} - + title = {Optimal brain damage}, + author = {LeCun, Yann and Denker, John and Solla, Sara}, + year = 1989, + journal = {Advances in neural information processing systems}, + volume = 2 +} @article{li2014communication, - author = {Li, Mu and Andersen, David G and Smola, Alexander J and Yu, Kai}, - journal = {Advances in Neural Information Processing Systems}, - title = {Communication efficient distributed machine learning with the parameter server}, - volume = 27, - year = 2014} - + title = {Communication efficient distributed machine learning with the parameter server}, + author = {Li, Mu and Andersen, David G and Smola, Alexander J and Yu, Kai}, + year = 2014, + journal = {Advances in Neural Information Processing Systems}, + volume = 27 +} @article{li2016lightrnn, - author = {Li, Xiang and Qin, Tao and Yang, Jian and Liu, Tie-Yan}, - journal = {Advances in Neural Information Processing Systems}, - title = {LightRNN: Memory and computation-efficient recurrent neural networks}, - volume = 29, - year = 2016} - + title = {LightRNN: Memory and computation-efficient recurrent neural networks}, + author = {Li, Xiang and Qin, Tao and Yang, Jian and Liu, Tie-Yan}, + year = 2016, + journal = {Advances in Neural Information Processing Systems}, + volume = 29 +} @article{li2017deep, - author = {Li, Yuxi}, - journal = {arXiv preprint arXiv:1701.07274}, - title = {Deep reinforcement learning: An overview}, - year = 2017} - + title = {Deep reinforcement learning: An overview}, + author = {Li, Yuxi}, + year = 2017, + journal = {arXiv preprint arXiv:1701.07274} +} @article{li2017learning, - author = {Li, Zhizhong and Hoiem, Derek}, - journal = {IEEE transactions on pattern analysis and machine intelligence}, - number = 12, - pages = {2935--2947}, - publisher = {IEEE}, - title = {Learning without forgetting}, - volume = 40, - year = 2017} - + title = {Learning without forgetting}, + author = {Li, Zhizhong and Hoiem, Derek}, + year = 2017, + journal = {IEEE transactions on pattern analysis and machine intelligence}, + publisher = {IEEE}, + volume = 40, + number = 12, + pages = {2935--2947} +} @article{li2019edge, - author = {Li, En and Zeng, Liekang and Zhou, Zhi and Chen, Xu}, - journal = {IEEE Transactions on Wireless Communications}, - number = 1, - pages = {447--457}, - publisher = {IEEE}, - title = {Edge AI: On-demand accelerating deep neural network inference via edge computing}, - volume = 19, - year = 2019} - + title = {Edge AI: On-demand accelerating deep neural network inference via edge computing}, + author = {Li, En and Zeng, Liekang and Zhou, Zhi and Chen, Xu}, + year = 2019, + journal = {IEEE Transactions on Wireless Communications}, + publisher = {IEEE}, + volume = 19, + number = 1, + pages = {447--457} +} +@inproceedings{Li2020Additive, + title = {Additive Powers-of-Two Quantization: An Efficient Non-uniform Discretization for Neural Networks}, + author = {Yuhang Li and Xin Dong and Wei Wang}, + year = 2020, + booktitle = {International Conference on Learning Representations}, + url = {https://openreview.net/forum?id=BkgXT24tDS}, + bdsk-url-1 = {https://openreview.net/forum?id=BkgXT24tDS} +} +@article{Li2020Federated, + title = {Federated Learning: Challenges, Methods, and Future Directions}, + author = {Li, Tian and Sahu, Anit Kumar and Talwalkar, Ameet and Smith, Virginia}, + year = 2020, + journal = {IEEE Signal Processing Magazine}, + volume = 37, + number = 3, + pages = {50--60}, + date-added = {2023-11-22 19:15:13 -0500}, + date-modified = {2023-11-22 19:17:19 -0500} +} @misc{liao_can_2023, - abstract = {Pruning is a widely used technique for reducing the size of deep neural networks while maintaining their performance. However, such a technique, despite being able to massively compress deep models, is hardly able to remove entire layers from a model (even when structured): is this an addressable task? In this study, we introduce EGP, an innovative Entropy Guided Pruning algorithm aimed at reducing the size of deep neural networks while preserving their performance. The key focus of EGP is to prioritize pruning connections in layers with low entropy, ultimately leading to their complete removal. Through extensive experiments conducted on popular models like ResNet-18 and Swin-T, our findings demonstrate that EGP effectively compresses deep neural networks while maintaining competitive performance levels. Our results not only shed light on the underlying mechanism behind the advantages of unstructured pruning, but also pave the way for further investigations into the intricate relationship between entropy, pruning techniques, and deep learning performance. The EGP algorithm and its insights hold great promise for advancing the field of network compression and optimization. The source code for EGP is released open-source.}, - author = {Liao, Zhu and Qu{\'e}tu, Victor and Nguyen, Van-Tam and Tartaglione, Enzo}, - doi = {10.48550/arXiv.2308.06619}, - file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/V6P3XB5H/Liao et al. - 2023 - Can Unstructured Pruning Reduce the Depth in Deep .pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/WSQ4ZUH4/2308.html:text/html}, - keywords = {Computer Science - Artificial Intelligence, Computer Science - Machine Learning}, - month = aug, - note = {arXiv:2308.06619 [cs]}, - publisher = {arXiv}, - title = {Can {Unstructured} {Pruning} {Reduce} the {Depth} in {Deep} {Neural} {Networks}?}, - url = {http://arxiv.org/abs/2308.06619}, - urldate = {2023-10-20}, - year = 2023, - Bdsk-Url-1 = {http://arxiv.org/abs/2308.06619}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.2308.06619}} - + title = {Can {Unstructured} {Pruning} {Reduce} the {Depth} in {Deep} {Neural} {Networks}?}, + author = {Liao, Zhu and Qu{\'e}tu, Victor and Nguyen, Van-Tam and Tartaglione, Enzo}, + year = 2023, + month = aug, + publisher = {arXiv}, + doi = {10.48550/arXiv.2308.06619}, + url = {http://arxiv.org/abs/2308.06619}, + urldate = {2023-10-20}, + note = {arXiv:2308.06619 [cs]}, + abstract = {Pruning is a widely used technique for reducing the size of deep neural networks while maintaining their performance. However, such a technique, despite being able to massively compress deep models, is hardly able to remove entire layers from a model (even when structured): is this an addressable task? In this study, we introduce EGP, an innovative Entropy Guided Pruning algorithm aimed at reducing the size of deep neural networks while preserving their performance. The key focus of EGP is to prioritize pruning connections in layers with low entropy, ultimately leading to their complete removal. Through extensive experiments conducted on popular models like ResNet-18 and Swin-T, our findings demonstrate that EGP effectively compresses deep neural networks while maintaining competitive performance levels. Our results not only shed light on the underlying mechanism behind the advantages of unstructured pruning, but also pave the way for further investigations into the intricate relationship between entropy, pruning techniques, and deep learning performance. The EGP algorithm and its insights hold great promise for advancing the field of network compression and optimization. The source code for EGP is released open-source.}, + file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/V6P3XB5H/Liao et al. - 2023 - Can Unstructured Pruning Reduce the Depth in Deep .pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/WSQ4ZUH4/2308.html:text/html}, + keywords = {Computer Science - Artificial Intelligence, Computer Science - Machine Learning}, + bdsk-url-1 = {http://arxiv.org/abs/2308.06619}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.2308.06619} +} +@misc{ligozat2022unraveling, + title = {Unraveling the Hidden Environmental Impacts of AI Solutions for Environment}, + author = {Anne-Laure Ligozat and Julien Lefèvre and Aurélie Bugeau and Jacques Combaz}, + year = 2022, + eprint = {2110.11822}, + archiveprefix = {arXiv}, + primaryclass = {cs.AI} +} @misc{lin_-device_2022, - annote = {Comment: NeurIPS 2022}, - author = {Lin, Ji and Zhu, Ligeng and Chen, Wei-Ming and Wang, Wei-Chen and Gan, Chuang and Han, Song}, - file = {Lin et al. - 2022 - On-Device Training Under 256KB Memory.pdf:/Users/alex/Zotero/storage/GMF6SWGT/Lin et al. - 2022 - On-Device Training Under 256KB Memory.pdf:application/pdf}, - keywords = {Computer Science - Computer Vision and Pattern Recognition}, - language = {en}, - month = nov, - note = {arXiv:2206.15472 [cs]}, - publisher = {arXiv}, - title = {On-{Device} {Training} {Under} {256KB} {Memory}}, - url = {http://arxiv.org/abs/2206.15472}, - urldate = {2023-10-26}, - year = 2022, - Bdsk-Url-1 = {http://arxiv.org/abs/2206.15472}} - + title = {On-{Device} {Training} {Under} {256KB} {Memory}}, + author = {Lin, Ji and Zhu, Ligeng and Chen, Wei-Ming and Wang, Wei-Chen and Gan, Chuang and Han, Song}, + year = 2022, + month = nov, + journal = {Advances in Neural Information Processing Systems}, + booktitle = {ArXiv}, + publisher = {arXiv}, + volume = 35, + pages = {22941--22954}, + url = {http://arxiv.org/abs/2206.15472}, + urldate = {2023-10-26}, + note = {arXiv:2206.15472 [cs]}, + annote = {Comment: NeurIPS 2022}, + file = {Lin et al. - 2022 - On-Device Training Under 256KB Memory.pdf:/Users/alex/Zotero/storage/GMF6SWGT/Lin et al. - 2022 - On-Device Training Under 256KB Memory.pdf:application/pdf}, + keywords = {Computer Science - Computer Vision and Pattern Recognition}, + language = {en}, + bdsk-url-1 = {http://arxiv.org/abs/2206.15472} +} @misc{lin_-device_2022-1, - annote = {Comment: NeurIPS 2022}, - author = {Lin, Ji and Zhu, Ligeng and Chen, Wei-Ming and Wang, Wei-Chen and Gan, Chuang and Han, Song}, - file = {Lin et al. - 2022 - On-Device Training Under 256KB Memory.pdf:/Users/alex/Zotero/storage/DNIY32R2/Lin et al. - 2022 - On-Device Training Under 256KB Memory.pdf:application/pdf}, - keywords = {Computer Science - Computer Vision and Pattern Recognition}, - language = {en}, - month = nov, - note = {arXiv:2206.15472 [cs]}, - publisher = {arXiv}, - title = {On-{Device} {Training} {Under} {256KB} {Memory}}, - url = {http://arxiv.org/abs/2206.15472}, - urldate = {2023-10-25}, - year = 2022, - Bdsk-Url-1 = {http://arxiv.org/abs/2206.15472}} - + title = {On-{Device} {Training} {Under} {256KB} {Memory}}, + author = {Lin, Ji and Zhu, Ligeng and Chen, Wei-Ming and Wang, Wei-Chen and Gan, Chuang and Han, Song}, + year = 2022, + month = nov, + publisher = {arXiv}, + url = {http://arxiv.org/abs/2206.15472}, + urldate = {2023-10-25}, + note = {arXiv:2206.15472 [cs]}, + language = {en}, + keywords = {Computer Science - Computer Vision and Pattern Recognition}, + annote = {Comment: NeurIPS 2022}, + file = {Lin et al. - 2022 - On-Device Training Under 256KB Memory.pdf:/Users/alex/Zotero/storage/DNIY32R2/Lin et al. - 2022 - On-Device Training Under 256KB Memory.pdf:application/pdf}, + bdsk-url-1 = {http://arxiv.org/abs/2206.15472} +} @misc{lin_mcunet_2020, - abstract = {Machine learning on tiny IoT devices based on microcontroller units (MCU) is appealing but challenging: the memory of microcontrollers is 2-3 orders of magnitude smaller even than mobile phones. We propose MCUNet, a framework that jointly designs the efficient neural architecture (TinyNAS) and the lightweight inference engine (TinyEngine), enabling ImageNet-scale inference on microcontrollers. TinyNAS adopts a two-stage neural architecture search approach that first optimizes the search space to fit the resource constraints, then specializes the network architecture in the optimized search space. TinyNAS can automatically handle diverse constraints (i.e.device, latency, energy, memory) under low search costs.TinyNAS is co-designed with TinyEngine, a memory-efficient inference library to expand the search space and fit a larger model. TinyEngine adapts the memory scheduling according to the overall network topology rather than layer-wise optimization, reducing the memory usage by 4.8x, and accelerating the inference by 1.7-3.3x compared to TF-Lite Micro and CMSIS-NN. MCUNet is the first to achieves {\textgreater}70\% ImageNet top1 accuracy on an off-the-shelf commercial microcontroller, using 3.5x less SRAM and 5.7x less Flash compared to quantized MobileNetV2 and ResNet-18. On visual\&audio wake words tasks, MCUNet achieves state-of-the-art accuracy and runs 2.4-3.4x faster than MobileNetV2 and ProxylessNAS-based solutions with 3.7-4.1x smaller peak SRAM. Our study suggests that the era of always-on tiny machine learning on IoT devices has arrived. Code and models can be found here: https://tinyml.mit.edu.}, - annote = {Comment: NeurIPS 2020 (spotlight)}, - author = {Lin, Ji and Chen, Wei-Ming and Lin, Yujun and Cohn, John and Gan, Chuang and Han, Song}, - doi = {10.48550/arXiv.2007.10319}, - file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/IX2JN4P9/Lin et al. - 2020 - MCUNet Tiny Deep Learning on IoT Devices.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/BAKHZ46Y/2007.html:text/html}, - keywords = {Computer Science - Computer Vision and Pattern Recognition}, - language = {en}, - month = nov, - note = {arXiv:2007.10319 [cs]}, - publisher = {arXiv}, - shorttitle = {{MCUNet}}, - title = {{MCUNet}: {Tiny} {Deep} {Learning} on {IoT} {Devices}}, - url = {http://arxiv.org/abs/2007.10319}, - urldate = {2023-10-20}, - year = 2020, - Bdsk-Url-1 = {http://arxiv.org/abs/2007.10319}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.2007.10319}} - + title = {{MCUNet}: {Tiny} {Deep} {Learning} on {IoT} {Devices}}, + shorttitle = {{MCUNet}}, + author = {Lin, Ji and Chen, Wei-Ming and Lin, Yujun and Cohn, John and Gan, Chuang and Han, Song}, + year = 2020, + month = nov, + journal = {Advances in Neural Information Processing Systems}, + publisher = {arXiv}, + volume = 33, + pages = {11711--11722}, + doi = {10.48550/arXiv.2007.10319}, + url = {http://arxiv.org/abs/2007.10319}, + urldate = {2023-10-20}, + note = {arXiv:2007.10319 [cs]}, + abstract = {Machine learning on tiny IoT devices based on microcontroller units (MCU) is appealing but challenging: the memory of microcontrollers is 2-3 orders of magnitude smaller even than mobile phones. We propose MCUNet, a framework that jointly designs the efficient neural architecture (TinyNAS) and the lightweight inference engine (TinyEngine), enabling ImageNet-scale inference on microcontrollers. TinyNAS adopts a two-stage neural architecture search approach that first optimizes the search space to fit the resource constraints, then specializes the network architecture in the optimized search space. TinyNAS can automatically handle diverse constraints (i.e.device, latency, energy, memory) under low search costs.TinyNAS is co-designed with TinyEngine, a memory-efficient inference library to expand the search space and fit a larger model. TinyEngine adapts the memory scheduling according to the overall network topology rather than layer-wise optimization, reducing the memory usage by 4.8x, and accelerating the inference by 1.7-3.3x compared to TF-Lite Micro and CMSIS-NN. MCUNet is the first to achieves {\textgreater}70\% ImageNet top1 accuracy on an off-the-shelf commercial microcontroller, using 3.5x less SRAM and 5.7x less Flash compared to quantized MobileNetV2 and ResNet-18. On visual\&audio wake words tasks, MCUNet achieves state-of-the-art accuracy and runs 2.4-3.4x faster than MobileNetV2 and ProxylessNAS-based solutions with 3.7-4.1x smaller peak SRAM. Our study suggests that the era of always-on tiny machine learning on IoT devices has arrived. Code and models can be found here: https://tinyml.mit.edu.}, + annote = {Comment: NeurIPS 2020 (spotlight)}, + file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/IX2JN4P9/Lin et al. - 2020 - MCUNet Tiny Deep Learning on IoT Devices.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/BAKHZ46Y/2007.html:text/html}, + keywords = {Computer Science - Computer Vision and Pattern Recognition}, + language = {en}, + bdsk-url-1 = {http://arxiv.org/abs/2007.10319}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.2007.10319}, + archiveprefix = {arXiv}, + eprint = {2007.10319}, + primaryclass = {cs.CV} +} @inproceedings{lin2014microsoft, - author = {Lin, Tsung-Yi and Maire, Michael and Belongie, Serge and Hays, James and Perona, Pietro and Ramanan, Deva and Doll{\'a}r, Piotr and Zitnick, C Lawrence}, - booktitle = {Computer Vision--ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part V 13}, - organization = {Springer}, - pages = {740--755}, - title = {Microsoft coco: Common objects in context}, - year = 2014} - + title = {Microsoft coco: Common objects in context}, + author = {Lin, Tsung-Yi and Maire, Michael and Belongie, Serge and Hays, James and Perona, Pietro and Ramanan, Deva and Doll{\'a}r, Piotr and Zitnick, C Lawrence}, + year = 2014, + booktitle = {Computer Vision--ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part V 13}, + pages = {740--755}, + organization = {Springer} +} @article{lin2020mcunet, + title = {Mcunet: Tiny deep learning on iot devices}, + author = {Lin, Ji and Chen, Wei-Ming and Lin, Yujun and Gan, Chuang and Han, Song and others}, + year = 2020, + journal = {Advances in Neural Information Processing Systems}, + volume = 33, + pages = {11711--11722}, + eprint = {2007.10319}, archiveprefix = {arXiv}, - author = {Lin, Ji and Chen, Wei-Ming and Lin, Yujun and Gan, Chuang and Han, Song and others}, - eprint = {2007.10319}, - journal = {Advances in Neural Information Processing Systems}, - pages = {11711--11722}, - primaryclass = {cs.CV}, - title = {Mcunet: Tiny deep learning on iot devices}, - volume = 33, - year = 2020} - + primaryclass = {cs.CV} +} @article{lin2022device, - author = {Lin, Ji and Zhu, Ligeng and Chen, Wei-Ming and Wang, Wei-Chen and Gan, Chuang and Han, Song}, - journal = {Advances in Neural Information Processing Systems}, - pages = {22941--22954}, - title = {On-device training under 256kb memory}, - volume = 35, - year = 2022} - -@misc{lu_notes_2016, - abstract = {Low-rank matrix factorization (MF) is an important technique in data science. The key idea of MF is that there exists latent structures in the data, by uncovering which we could obtain a compressed representation of the data. By factorizing an original matrix to low-rank matrices, MF provides a unified method for dimension reduction, clustering, and matrix completion. In this article we review several important variants of MF, including: Basic MF, Non-negative MF, Orthogonal non-negative MF. As can be told from their names, non-negative MF and orthogonal non-negative MF are variants of basic MF with non-negativity and/or orthogonality constraints. Such constraints are useful in specific senarios. In the first part of this article, we introduce, for each of these models, the application scenarios, the distinctive properties, and the optimizing method. By properly adapting MF, we can go beyond the problem of clustering and matrix completion. In the second part of this article, we will extend MF to sparse matrix compeletion, enhance matrix compeletion using various regularization methods, and make use of MF for (semi-)supervised learning by introducing latent space reinforcement and transformation. We will see that MF is not only a useful model but also as a flexible framework that is applicable for various prediction problems.}, - author = {Lu, Yuan and Yang, Jie}, - doi = {10.48550/arXiv.1507.00333}, - file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/4QED5ZU9/Lu and Yang - 2016 - Notes on Low-rank Matrix Factorization.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/XIBZBDJQ/1507.html:text/html}, - keywords = {Computer Science - Information Retrieval, Computer Science - Machine Learning, Mathematics - Numerical Analysis}, - month = may, - note = {arXiv:1507.00333 [cs]}, - publisher = {arXiv}, - title = {Notes on {Low}-rank {Matrix} {Factorization}}, - url = {http://arxiv.org/abs/1507.00333}, - urldate = {2023-10-20}, - year = 2016, - Bdsk-Url-1 = {http://arxiv.org/abs/1507.00333}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.1507.00333}} - -@article{lundberg2017unified, - author = {Lundberg, Scott M and Lee, Su-In}, - journal = {Advances in neural information processing systems}, - title = {A unified approach to interpreting model predictions}, - volume = 30, - year = 2017} - -@article{mattson2020mlperf, - author = {Mattson, Peter and Cheng, Christine and Diamos, Gregory and Coleman, Cody and Micikevicius, Paulius and Patterson, David and Tang, Hanlin and Wei, Gu-Yeon and Bailis, Peter and Bittorf, Victor and others}, - journal = {Proceedings of Machine Learning and Systems}, - pages = {336--349}, - title = {Mlperf training benchmark}, - volume = 2, - year = 2020} - + title = {On-device training under 256kb memory}, + author = {Lin, Ji and Zhu, Ligeng and Chen, Wei-Ming and Wang, Wei-Chen and Gan, Chuang and Han, Song}, + year = 2022, + journal = {Advances in Neural Information Processing Systems}, + volume = 35, + pages = {22941--22954} +} +@inproceedings{lin2022ondevice, + title = {On-Device Training Under 256KB Memory}, + author = {Lin, Ji and Zhu, Ligeng and Chen, Wei-Ming and Wang, Wei-Chen and Gan, Chuang and Han, Song}, + year = 2022, + booktitle = {ArXiv} +} +@article{lin2023awq, + title = {AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration}, + author = {Lin, Ji and Tang, Jiaming and Tang, Haotian and Yang, Shang and Dang, Xingyu and Han, Song}, + year = 2023, + journal = {arXiv} +} +@book{lindgren2023handbook, + title = {Handbook of Critical Studies of Artificial Intelligence}, + author = {Lindgren, Simon}, + year = 2023, + publisher = {Edward Elgar Publishing} +} +@article{lindholm_nvidia_2008, + title = {{NVIDIA} {Tesla}: {A} {Unified} {Graphics} and {Computing} {Architecture}}, + shorttitle = {{NVIDIA} {Tesla}}, + author = {Lindholm, Erik and Nickolls, John and Oberman, Stuart and Montrym, John}, + year = 2008, + month = mar, + journal = {IEEE Micro}, + volume = 28, + number = 2, + pages = {39--55}, + doi = {10.1109/MM.2008.31}, + issn = {1937-4143}, + url = {https://ieeexplore.ieee.org/document/4523358}, + urldate = {2023-11-07}, + note = {Conference Name: IEEE Micro}, + abstract = {To enable flexible, programmable graphics and high-performance computing, NVIDIA has developed the Tesla scalable unified graphics and parallel computing architecture. Its scalable parallel array of processors is massively multithreaded and programmable in C or via graphics APIs.}, + bdsk-url-1 = {https://ieeexplore.ieee.org/document/4523358}, + bdsk-url-2 = {https://doi.org/10.1109/MM.2008.31} +} +@inproceedings{Lipp2018meltdown, + title = {Meltdown: Reading Kernel Memory from User Space}, + author = {Moritz Lipp and Michael Schwarz and Daniel Gruss and Thomas Prescher and Werner Haas and Anders Fogh and Jann Horn and Stefan Mangard and Paul Kocher and Daniel Genkin and Yuval Yarom and Mike Hamburg}, + year = 2018, + booktitle = {27th {USENIX} Security Symposium ({USENIX} Security 18)}, + date-added = {2023-11-22 16:32:26 -0500}, + date-modified = {2023-11-22 16:33:08 -0500} +} +@article{loh20083d, + title = {3D-stacked memory architectures for multi-core processors}, + author = {Loh, Gabriel H}, + year = 2008, + journal = {ACM SIGARCH computer architecture news}, + publisher = {ACM New York, NY, USA}, + volume = 36, + number = 3, + pages = {453--464} +} +@inproceedings{long2020ai, + title = {What is AI literacy? Competencies and design considerations}, + author = {Long, Duri and Magerko, Brian}, + year = 2020, + booktitle = {Proceedings of the 2020 CHI conference on human factors in computing systems}, + pages = {1--16} +} +@inproceedings{lou2013accurate, + title = {Accurate intelligible models with pairwise interactions}, + author = {Lou, Yin and Caruana, Rich and Gehrke, Johannes and Hooker, Giles}, + year = 2013, + booktitle = {Proceedings of the 19th ACM SIGKDD international conference on Knowledge discovery and data mining}, + pages = {623--631} +} +@article{lowy2021fermi, + title = {Fermi: Fair empirical risk minimization via exponential R{\'e}nyi mutual information}, + author = {Lowy, Andrew and Pavan, Rakesh and Baharlouei, Sina and Razaviyayn, Meisam and Beirami, Ahmad}, + year = 2021 +} +@misc{lu_notes_2016, + title = {Notes on {Low}-rank {Matrix} {Factorization}}, + author = {Lu, Yuan and Yang, Jie}, + year = 2016, + month = may, + publisher = {arXiv}, + doi = {10.48550/arXiv.1507.00333}, + url = {http://arxiv.org/abs/1507.00333}, + urldate = {2023-10-20}, + note = {arXiv:1507.00333 [cs]}, + abstract = {Low-rank matrix factorization (MF) is an important technique in data science. The key idea of MF is that there exists latent structures in the data, by uncovering which we could obtain a compressed representation of the data. By factorizing an original matrix to low-rank matrices, MF provides a unified method for dimension reduction, clustering, and matrix completion. In this article we review several important variants of MF, including: Basic MF, Non-negative MF, Orthogonal non-negative MF. As can be told from their names, non-negative MF and orthogonal non-negative MF are variants of basic MF with non-negativity and/or orthogonality constraints. Such constraints are useful in specific senarios. In the first part of this article, we introduce, for each of these models, the application scenarios, the distinctive properties, and the optimizing method. By properly adapting MF, we can go beyond the problem of clustering and matrix completion. In the second part of this article, we will extend MF to sparse matrix compeletion, enhance matrix compeletion using various regularization methods, and make use of MF for (semi-)supervised learning by introducing latent space reinforcement and transformation. We will see that MF is not only a useful model but also as a flexible framework that is applicable for various prediction problems.}, + file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/4QED5ZU9/Lu and Yang - 2016 - Notes on Low-rank Matrix Factorization.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/XIBZBDJQ/1507.html:text/html}, + keywords = {Computer Science - Information Retrieval, Computer Science - Machine Learning, Mathematics - Numerical Analysis}, + bdsk-url-1 = {http://arxiv.org/abs/1507.00333}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.1507.00333} +} +@inproceedings{luebke2008cuda, + title = {CUDA: Scalable parallel programming for high-performance scientific computing}, + author = {Luebke, David}, + year = 2008, + booktitle = {2008 5th IEEE International Symposium on Biomedical Imaging: From Nano to Macro}, + volume = {}, + number = {}, + pages = {836--838}, + doi = {10.1109/ISBI.2008.4541126}, + bdsk-url-1 = {https://doi.org/10.1109/ISBI.2008.4541126} +} +@article{lundberg2017unified, + title = {A unified approach to interpreting model predictions}, + author = {Lundberg, Scott M and Lee, Su-In}, + year = 2017, + journal = {Advances in neural information processing systems}, + volume = 30 +} +@article{maass1997networks, + title = {Networks of spiking neurons: the third generation of neural network models}, + author = {Maass, Wolfgang}, + year = 1997, + journal = {Neural networks}, + publisher = {Elsevier}, + volume = 10, + number = 9, + pages = {1659--1671} +} +@article{MAL-083, + title = {Advances and Open Problems in Federated Learning}, + author = {Peter Kairouz and H. Brendan McMahan and Brendan Avent and Aur{\'e}lien Bellet and Mehdi Bennis and Arjun Nitin Bhagoji and Kallista Bonawitz and Zachary Charles and Graham Cormode and Rachel Cummings and Rafael G. L. D'Oliveira and Hubert Eichner and Salim El Rouayheb and David Evans and Josh Gardner and Zachary Garrett and Adri{\`a} Gasc{\'o}n and Badih Ghazi and Phillip B. Gibbons and Marco Gruteser and Zaid Harchaoui and Chaoyang He and Lie He and Zhouyuan Huo and Ben Hutchinson and Justin Hsu and Martin Jaggi and Tara Javidi and Gauri Joshi and Mikhail Khodak and Jakub Konecn{\'y} and Aleksandra Korolova and Farinaz Koushanfar and Sanmi Koyejo and Tancr{\`e}de Lepoint and Yang Liu and Prateek Mittal and Mehryar Mohri and Richard Nock and Ayfer {\"O}zg{\"u}r and Rasmus Pagh and Hang Qi and Daniel Ramage and Ramesh Raskar and Mariana Raykova and Dawn Song and Weikang Song and Sebastian U. Stich and Ziteng Sun and Ananda Theertha Suresh and Florian Tram{\`e}r and Praneeth Vepakomma and Jianyu Wang and Li Xiong and Zheng Xu and Qiang Yang and Felix X. Yu and Han Yu and Sen Zhao}, + year = 2021, + journal = {Foundations and Trends{\textregistered} in Machine Learning}, + volume = 14, + number = {1--2}, + pages = {1--210}, + doi = {10.1561/2200000083}, + issn = {1935-8237}, + url = {http://dx.doi.org/10.1561/2200000083}, + date-added = {2023-11-22 19:14:08 -0500}, + date-modified = {2023-11-22 19:14:08 -0500}, + bdsk-url-1 = {http://dx.doi.org/10.1561/2200000083} +} +@article{markovic2020, + title = {Physics for neuromorphic computing}, + author = {Markovi{\'c}, Danijela and Mizrahi, Alice and Querlioz, Damien and Grollier, Julie}, + year = 2020, + journal = {Nature Reviews Physics}, + publisher = {Nature Publishing Group UK London}, + volume = 2, + number = 9, + pages = {499--510} +} +@article{martin1993myth, + title = {The myth of the awesome thinking machine}, + author = {Martin, C Dianne}, + year = 1993, + journal = {Communications of the ACM}, + publisher = {ACM New York, NY, USA}, + volume = 36, + number = 4, + pages = {120--133} +} +@article{maslej2023artificial, + title = {Artificial intelligence index report 2023}, + author = {Maslej, Nestor and Fattorini, Loredana and Brynjolfsson, Erik and Etchemendy, John and Ligett, Katrina and Lyons, Terah and Manyika, James and Ngo, Helen and Niebles, Juan Carlos and Parli, Vanessa and others}, + year = 2023, + journal = {arXiv preprint arXiv:2310.03715} +} +@article{mattson2020mlperf, + title = {Mlperf training benchmark}, + author = {Mattson, Peter and Cheng, Christine and Diamos, Gregory and Coleman, Cody and Micikevicius, Paulius and Patterson, David and Tang, Hanlin and Wei, Gu-Yeon and Bailis, Peter and Bittorf, Victor and others}, + year = 2020, + journal = {Proceedings of Machine Learning and Systems}, + volume = 2, + pages = {336--349} +} +@article{maxime2016impact, + title = {The Impact of Demand Uncertainty on Consumer Subsidies for Green Technology Adoption}, + author = {Maxime C. Cohen and Ruben Lobel and Georgia Perakis}, + year = 2016, + journal = {Management Science}, + publisher = {INFORMS}, + volume = 62, + number = 5, + pages = {1235--1258}, + issn = {00251909, 15265501}, + url = {http://www.jstor.org/stable/43835073}, + urldate = {2023-12-01}, + abstract = {This paper studies government subsidies for green technology adoption while considering the manufacturing industry's response. Government subsidies offered directly to consumers impact the supplier's production and pricing decisions. Our analysis expands the current understanding of the price-setting newsvendor model, incorporating the external influence from the government, who is now an additional player in the system. We quantify how demand uncertainty impacts the various players (government, industry, and consumers) when designing policies. We further show that, for convex demand functions, an increase in demand uncertainty leads to higher production quantities and lower prices, resulting in lower profits for the supplier. With this in mind, one could expect consumer surplus to increase with uncertainty. In fact, we show that this is not always the case and that the uncertainty impact on consumer surplus depends on the trade-off between lower prices and the possibility of underserving customers with high valuations. We also show that when policy makers such as governments ignore demand uncertainty when designing consumer subsidies, they can significantly miss the desired adoption target level. From a coordination perspective, we demonstrate that the decentralized decisions are also optimal for a central planner managing jointly the supplier and the government. As a result, subsidies provide a coordination mechanism.} +} +@incollection{mccarthy1981epistemological, + title = {Epistemological problems of artificial intelligence}, + author = {McCarthy, John}, + year = 1981, + booktitle = {Readings in artificial intelligence}, + publisher = {Elsevier}, + pages = {459--465} +} @inproceedings{mcmahan2017communication, - author = {McMahan, Brendan and Moore, Eider and Ramage, Daniel and Hampson, Seth and y Arcas, Blaise Aguera}, - booktitle = {Artificial intelligence and statistics}, - organization = {PMLR}, - pages = {1273--1282}, - title = {Communication-efficient learning of deep networks from decentralized data}, - year = 2017} - + title = {Communication-efficient learning of deep networks from decentralized data}, + author = {McMahan, Brendan and Moore, Eider and Ramage, Daniel and Hampson, Seth and y Arcas, Blaise Aguera}, + year = 2017, + booktitle = {Artificial intelligence and statistics}, + pages = {1273--1282}, + organization = {PMLR} +} @inproceedings{mcmahan2023communicationefficient, - author = {McMahan, Brendan and Moore, Eider and Ramage, Daniel and Hampson, Seth and y Arcas, Blaise Aguera}, - booktitle = {Artificial intelligence and statistics}, - organization = {PMLR}, - pages = {1273--1282}, - title = {Communication-efficient learning of deep networks from decentralized data}, - year = 2017} - + title = {Communication-efficient learning of deep networks from decentralized data}, + author = {McMahan, Brendan and Moore, Eider and Ramage, Daniel and Hampson, Seth and y Arcas, Blaise Aguera}, + year = 2017, + booktitle = {Artificial intelligence and statistics}, + pages = {1273--1282}, + organization = {PMLR} +} +@article{miller2000optical, + title = {Optical interconnects to silicon}, + author = {Miller, David AB}, + year = 2000, + journal = {IEEE Journal of Selected Topics in Quantum Electronics}, + publisher = {IEEE}, + volume = 6, + number = 6, + pages = {1312--1317} +} +@article{miller2015remote, + title = {Remote exploitation of an unaltered passenger vehicle}, + author = {Miller, Charlie and Valasek, Chris}, + year = 2015, + journal = {Black Hat USA}, + volume = 2015, + number = {S 91}, + pages = {1--91}, + date-added = {2023-11-22 17:11:27 -0500}, + date-modified = {2023-11-22 17:12:18 -0500} +} +@article{miller2019lessons, + title = {Lessons learned from hacking a car}, + author = {Miller, Charlie}, + year = 2019, + journal = {IEEE Design & Test}, + volume = 36, + number = 6, + pages = {7--9}, + date-added = {2023-11-22 16:12:04 -0500}, + date-modified = {2023-11-22 16:13:31 -0500} +} +@article{mills1997overview, + title = {An overview of semiconductor photocatalysis}, + author = {Mills, Andrew and Le Hunte, Stephen}, + year = 1997, + journal = {Journal of photochemistry and photobiology A: Chemistry}, + publisher = {Elsevier}, + volume = 108, + number = 1, + pages = {1--35} +} +@article{mittal2021survey, + title = {A survey of SRAM-based in-memory computing techniques and applications}, + author = {Mittal, Sparsh and Verma, Gaurav and Kaushik, Brajesh and Khanday, Farooq A}, + year = 2021, + journal = {Journal of Systems Architecture}, + publisher = {Elsevier}, + volume = 119, + pages = 102276 +} +@article{modha2023neural, + title = {Neural inference at the frontier of energy, space, and time}, + author = {Modha, Dharmendra S and Akopyan, Filipp and Andreopoulos, Alexander and Appuswamy, Rathinakumar and Arthur, John V and Cassidy, Andrew S and Datta, Pallab and DeBole, Michael V and Esser, Steven K and Otero, Carlos Ortega and others}, + year = 2023, + journal = {Science}, + publisher = {American Association for the Advancement of Science}, + volume = 382, + number = 6668, + pages = {329--335} +} +@article{monyei2018electrons, + title = {Electrons have no identity: setting right misrepresentations in Google and Apple’s clean energy purchasing}, + author = {Monyei, Chukwuka G and Jenkins, Kirsten EH}, + year = 2018, + journal = {Energy research \& social science}, + publisher = {Elsevier}, + volume = 46, + pages = {48--51} +} @article{moshawrab2023reviewing, - author = {Moshawrab, Mohammad and Adda, Mehdi and Bouzouane, Abdenour and Ibrahim, Hussein and Raad, Ali}, - journal = {Electronics}, - number = 10, - pages = 2287, - publisher = {MDPI}, - title = {Reviewing Federated Learning Aggregation Algorithms; Strategies, Contributions, Limitations and Future Perspectives}, - volume = 12, - year = 2023} - + title = {Reviewing Federated Learning Aggregation Algorithms; Strategies, Contributions, Limitations and Future Perspectives}, + author = {Moshawrab, Mohammad and Adda, Mehdi and Bouzouane, Abdenour and Ibrahim, Hussein and Raad, Ali}, + year = 2023, + journal = {Electronics}, + publisher = {MDPI}, + volume = 12, + number = 10, + pages = 2287 +} +@article{muhammad2022survey, + title = {A survey on efficient methods for adversarial robustness}, + author = {Muhammad, Awais and Bae, Sung-Ho}, + year = 2022, + journal = {IEEE Access}, + publisher = {IEEE}, + volume = 10, + pages = {118815--118830} +} +@inproceedings{munshi2009opencl, + title = {The OpenCL specification}, + author = {Munshi, Aaftab}, + year = 2009, + booktitle = {2009 IEEE Hot Chips 21 Symposium (HCS)}, + volume = {}, + number = {}, + pages = {1--314}, + doi = {10.1109/HOTCHIPS.2009.7478342}, + bdsk-url-1 = {https://doi.org/10.1109/HOTCHIPS.2009.7478342} +} +@article{musk2019integrated, + title = {An integrated brain-machine interface platform with thousands of channels}, + author = {Musk, Elon and others}, + year = 2019, + journal = {Journal of medical Internet research}, + publisher = {JMIR Publications Inc., Toronto, Canada}, + volume = 21, + number = 10, + pages = {e16194} +} +@book{nakano2021geopolitics, + title = {The geopolitics of critical minerals supply chains}, + author = {Nakano, Jane}, + year = 2021, + publisher = {JSTOR} +} +@article{narayanan2006break, + title = {How to break anonymity of the netflix prize dataset}, + author = {Narayanan, Arvind and Shmatikov, Vitaly}, + year = 2006, + journal = {arXiv preprint cs/0610105}, + date-added = {2023-11-22 16:16:19 -0500}, + date-modified = {2023-11-22 16:16:59 -0500} +} +@misc{nas, + title = {Neural Architecture Search with Reinforcement Learning}, + author = {Barret Zoph and Quoc V. Le}, + year = 2017, + eprint = {1611.01578}, + archiveprefix = {arXiv}, + primaryclass = {cs.LG} +} +@article{ng2021ai, + title = {AI literacy: Definition, teaching, evaluation and ethical issues}, + author = {Ng, Davy Tsz Kit and Leung, Jac Ka Lok and Chu, Kai Wah Samuel and Qiao, Maggie Shen}, + year = 2021, + journal = {Proceedings of the Association for Information Science and Technology}, + publisher = {Wiley Online Library}, + volume = 58, + number = 1, + pages = {504--509} +} +@article{ngo2022alignment, + title = {The alignment problem from a deep learning perspective}, + author = {Ngo, Richard and Chan, Lawrence and Mindermann, S{\"o}ren}, + year = 2022, + journal = {arXiv preprint arXiv:2209.00626} +} @inproceedings{nguyen2023re, - author = {Nguyen, Ngoc-Bao and Chandrasegaran, Keshigeyan and Abdollahzadeh, Milad and Cheung, Ngai-Man}, - booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition}, - pages = {16384--16393}, - title = {Re-thinking Model Inversion Attacks Against Deep Neural Networks}, - year = 2023} - + title = {Re-thinking Model Inversion Attacks Against Deep Neural Networks}, + author = {Nguyen, Ngoc-Bao and Chandrasegaran, Keshigeyan and Abdollahzadeh, Milad and Cheung, Ngai-Man}, + year = 2023, + booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition}, + pages = {16384--16393} +} +@misc{noauthor_amd_nodate, + title = {{AMD} {Radeon} {RX} 7000 {Series} {Desktop} {Graphics} {Cards}}, + url = {https://www.amd.com/en/graphics/radeon-rx-graphics}, + urldate = {2023-11-07}, + bdsk-url-1 = {https://www.amd.com/en/graphics/radeon-rx-graphics} +} @misc{noauthor_deep_nodate, - author = {Ivy Gu}, - title = {Deep {Learning} {Model} {Compression} (ii) {\textbar} by {Ivy} {Gu} {\textbar} {Medium}}, - url = {https://ivygdy.medium.com/deep-learning-model-compression-ii-546352ea9453}, - urldate = {2023-10-20}, - year = {2023}, - Bdsk-Url-1 = {https://ivygdy.medium.com/deep-learning-model-compression-ii-546352ea9453}} - + title = {Deep {Learning} {Model} {Compression} (ii) {\textbar} by {Ivy} {Gu} {\textbar} {Medium}}, + author = {Ivy Gu}, + year = 2023, + url = {https://ivygdy.medium.com/deep-learning-model-compression-ii-546352ea9453}, + urldate = {2023-10-20}, + bdsk-url-1 = {https://ivygdy.medium.com/deep-learning-model-compression-ii-546352ea9453} +} +@misc{noauthor_evolution_2023, + title = {The {Evolution} of {Audio} {DSPs}}, + year = 2023, + month = oct, + journal = {audioXpress}, + url = {https://audioxpress.com/article/the-evolution-of-audio-dsps}, + urldate = {2023-11-07}, + abstract = {To complement the extensive perspective of another Market Update feature article on DSP Products and Applications, published in the November 2020 edition, audioXpress was honored to have the valuable contribution from one of the main suppliers in the field. In this article, Youval Nachum, CEVA's Senior Product Marketing Manager, writes about \"The Evolution of Audio DSPs,\" discussing how DSP technology has evolved, its impact on the user experience, and what the future of DSP has in store for us.}, + language = {en}, + bdsk-url-1 = {https://audioxpress.com/article/the-evolution-of-audio-dsps} +} +@misc{noauthor_fpga_nodate, + title = {{FPGA} {Architecture} {Overview}}, + url = {https://www.intel.com/content/www/us/en/docs/oneapi-fpga-add-on/optimization-guide/2023-1/fpga-architecture-overview.html}, + urldate = {2023-11-07}, + bdsk-url-1 = {https://www.intel.com/content/www/us/en/docs/oneapi-fpga-add-on/optimization-guide/2023-1/fpga-architecture-overview.html} +} +@misc{noauthor_google_2023, + title = {Google {Tensor} {G3}: {The} new chip that gives your {Pixel} an {AI} upgrade}, + shorttitle = {Google {Tensor} {G3}}, + year = 2023, + month = oct, + journal = {Google}, + url = {https://blog.google/products/pixel/google-tensor-g3-pixel-8/}, + urldate = {2023-11-07}, + abstract = {Tensor G3 on Pixel 8 and Pixel 8 Pro is more helpful, more efficient and more powerful.}, + language = {en-us}, + bdsk-url-1 = {https://blog.google/products/pixel/google-tensor-g3-pixel-8/} +} +@misc{noauthor_hexagon_nodate, + title = {Hexagon {DSP} {SDK} {Processor}}, + journal = {Qualcomm Developer Network}, + url = {https://developer.qualcomm.com/software/hexagon-dsp-sdk/dsp-processor}, + urldate = {2023-11-07}, + abstract = {The Hexagon DSP processor has both CPU and DSP functionality to support deeply embedded processing needs of the mobile platform for both multimedia and modem functions.}, + language = {en}, + bdsk-url-1 = {https://developer.qualcomm.com/software/hexagon-dsp-sdk/dsp-processor} +} +@misc{noauthor_integrated_2023, + title = {Integrated circuit}, + year = 2023, + month = nov, + journal = {Wikipedia}, + url = {https://en.wikipedia.org/w/index.php?title=Integrated_circuit&oldid=1183537457}, + urldate = {2023-11-07}, + copyright = {Creative Commons Attribution-ShareAlike License}, + note = {Page Version ID: 1183537457}, + abstract = {An integrated circuit (also known as an IC, a chip, or a microchip) is a set of electronic circuits on one small flat piece of semiconductor material, usually silicon. Large numbers of miniaturized transistors and other electronic components are integrated together on the chip. This results in circuits that are orders of magnitude smaller, faster, and less expensive than those constructed of discrete components, allowing a large transistor count. The IC's mass production capability, reliability, and building-block approach to integrated circuit design have ensured the rapid adoption of standardized ICs in place of designs using discrete transistors. ICs are now used in virtually all electronic equipment and have revolutionized the world of electronics. Computers, mobile phones and other home appliances are now essential parts of the structure of modern societies, made possible by the small size and low cost of ICs such as modern computer processors and microcontrollers. Very-large-scale integration was made practical by technological advancements in semiconductor device fabrication. Since their origins in the 1960s, the size, speed, and capacity of chips have progressed enormously, driven by technical advances that fit more and more transistors on chips of the same size -- a modern chip may have many billions of transistors in an area the size of a human fingernail. These advances, roughly following Moore's law, make the computer chips of today possess millions of times the capacity and thousands of times the speed of the computer chips of the early 1970s. ICs have three main advantages over discrete circuits: size, cost and performance. The size and cost is low because the chips, with all their components, are printed as a unit by photolithography rather than being constructed one transistor at a time. Furthermore, packaged ICs use much less material than discrete circuits. Performance is high because the IC's components switch quickly and consume comparatively little power because of their small size and proximity. The main disadvantage of ICs is the high initial cost of designing them and the enormous capital cost of factory construction. This high initial cost means ICs are only commercially viable when high production volumes are anticipated.}, + language = {en}, + bdsk-url-1 = {https://en.wikipedia.org/w/index.php?title=Integrated_circuit&oldid=1183537457} +} +@misc{noauthor_intel_nodate, + title = {Intel{\textregistered} {Arc}{\texttrademark} {Graphics} {Overview}}, + journal = {Intel}, + url = {https://www.intel.com/content/www/us/en/products/details/discrete-gpus/arc.html}, + urldate = {2023-11-07}, + abstract = {Find out how Intel{\textregistered} Arc Graphics unlock lifelike gaming and seamless content creation.}, + language = {en}, + bdsk-url-1 = {https://www.intel.com/content/www/us/en/products/details/discrete-gpus/arc.html} +} @misc{noauthor_introduction_nodate, - author = {Hegde, Sumant}, - title = {An {Introduction} to {Separable} {Convolutions} - {Analytics} {Vidhya}}, - url = {https://www.analyticsvidhya.com/blog/2021/11/an-introduction-to-separable-convolutions/}, - urldate = {2023-10-20}, - year = {2023}, - Bdsk-Url-1 = {https://www.analyticsvidhya.com/blog/2021/11/an-introduction-to-separable-convolutions/}} - + title = {An {Introduction} to {Separable} {Convolutions} - {Analytics} {Vidhya}}, + author = {Hegde, Sumant}, + year = 2023, + url = {https://www.analyticsvidhya.com/blog/2021/11/an-introduction-to-separable-convolutions/}, + urldate = {2023-10-20}, + bdsk-url-1 = {https://www.analyticsvidhya.com/blog/2021/11/an-introduction-to-separable-convolutions/} +} @misc{noauthor_knowledge_nodate, - author = {IntelLabs}, - title = {Knowledge {Distillation} - {Neural} {Network} {Distiller}}, - url = {https://intellabs.github.io/distiller/knowledge_distillation.html}, - urldate = {2023-10-20}, - year = {2023}, - Bdsk-Url-1 = {https://intellabs.github.io/distiller/knowledge_distillation.html}} - + title = {Knowledge {Distillation} - {Neural} {Network} {Distiller}}, + author = {IntelLabs}, + year = 2023, + url = {https://intellabs.github.io/distiller/knowledge_distillation.html}, + urldate = {2023-10-20}, + bdsk-url-1 = {https://intellabs.github.io/distiller/knowledge_distillation.html} +} +@misc{noauthor_project_nodate, + title = {Project {Catapult} - {Microsoft} {Research}}, + url = {https://www.microsoft.com/en-us/research/project/project-catapult/}, + urldate = {2023-11-07}, + bdsk-url-1 = {https://www.microsoft.com/en-us/research/project/project-catapult/} +} +@misc{noauthor_what_nodate, + title = {What is an {FPGA}? {Field} {Programmable} {Gate} {Array}}, + shorttitle = {What is an {FPGA}?}, + journal = {AMD}, + url = {https://www.xilinx.com/products/silicon-devices/fpga/what-is-an-fpga.html}, + urldate = {2023-11-07}, + abstract = {What is an FPGA - Field Programmable Gate Arrays are semiconductor devices that are based around a matrix of configurable logic blocks (CLBs) connected via programmable interconnects. FPGAs can be reprogrammed to desired application or functionality requirements after manufacturing.}, + language = {en}, + bdsk-url-1 = {https://www.xilinx.com/products/silicon-devices/fpga/what-is-an-fpga.html} +} +@misc{noauthor_who_nodate, + title = {Who {Invented} the {Microprocessor}? - {CHM}}, + url = {https://computerhistory.org/blog/who-invented-the-microprocessor/}, + urldate = {2023-11-07}, + bdsk-url-1 = {https://computerhistory.org/blog/who-invented-the-microprocessor/} +} +@inproceedings{Norman2017TPUv1, + title = {In-Datacenter Performance Analysis of a Tensor Processing Unit}, + author = {Jouppi, Norman P. and Young, Cliff and Patil, Nishant and Patterson, David and Agrawal, Gaurav and Bajwa, Raminder and Bates, Sarah and Bhatia, Suresh and Boden, Nan and Borchers, Al and Boyle, Rick and Cantin, Pierre-luc and Chao, Clifford and Clark, Chris and Coriell, Jeremy and Daley, Mike and Dau, Matt and Dean, Jeffrey and Gelb, Ben and Ghaemmaghami, Tara Vazir and Gottipati, Rajendra and Gulland, William and Hagmann, Robert and Ho, C. Richard and Hogberg, Doug and Hu, John and Hundt, Robert and Hurt, Dan and Ibarz, Julian and Jaffey, Aaron and Jaworski, Alek and Kaplan, Alexander and Khaitan, Harshit and Killebrew, Daniel and Koch, Andy and Kumar, Naveen and Lacy, Steve and Laudon, James and Law, James and Le, Diemthu and Leary, Chris and Liu, Zhuyuan and Lucke, Kyle and Lundin, Alan and MacKean, Gordon and Maggiore, Adriana and Mahony, Maire and Miller, Kieran and Nagarajan, Rahul and Narayanaswami, Ravi and Ni, Ray and Nix, Kathy and Norrie, Thomas and Omernick, Mark and Penukonda, Narayana and Phelps, Andy and Ross, Jonathan and Ross, Matt and Salek, Amir and Samadiani, Emad and Severn, Chris and Sizikov, Gregory and Snelham, Matthew and Souter, Jed and Steinberg, Dan and Swing, Andy and Tan, Mercedes and Thorson, Gregory and Tian, Bo and Toma, Horia and Tuttle, Erick and Vasudevan, Vijay and Walter, Richard and Wang, Walter and Wilcox, Eric and Yoon, Doe Hyun}, + year = 2017, + booktitle = {Proceedings of the 44th Annual International Symposium on Computer Architecture}, + location = {Toronto, ON, Canada}, + publisher = {Association for Computing Machinery}, + address = {New York, NY, USA}, + series = {ISCA '17}, + pages = {1--12}, + doi = {10.1145/3079856.3080246}, + isbn = 9781450348928, + url = {https://doi.org/10.1145/3079856.3080246}, + abstract = {Many architects believe that major improvements in cost-energy-performance must now come from domain-specific hardware. This paper evaluates a custom ASIC---called a Tensor Processing Unit (TPU) --- deployed in datacenters since 2015 that accelerates the inference phase of neural networks (NN). The heart of the TPU is a 65,536 8-bit MAC matrix multiply unit that offers a peak throughput of 92 TeraOps/second (TOPS) and a large (28 MiB) software-managed on-chip memory. The TPU's deterministic execution model is a better match to the 99th-percentile response-time requirement of our NN applications than are the time-varying optimizations of CPUs and GPUs that help average throughput more than guaranteed latency. The lack of such features helps explain why, despite having myriad MACs and a big memory, the TPU is relatively small and low power. We compare the TPU to a server-class Intel Haswell CPU and an Nvidia K80 GPU, which are contemporaries deployed in the same datacenters. Our workload, written in the high-level TensorFlow framework, uses production NN applications (MLPs, CNNs, and LSTMs) that represent 95\% of our datacenters' NN inference demand. Despite low utilization for some applications, the TPU is on average about 15X -- 30X faster than its contemporary GPU or CPU, with TOPS/Watt about 30X -- 80X higher. Moreover, using the CPU's GDDR5 memory in the TPU would triple achieved TOPS and raise TOPS/Watt to nearly 70X the GPU and 200X the CPU.}, + numpages = 12, + keywords = {accelerator, neural network, MLP, TPU, CNN, deep learning, domain-specific architecture, GPU, TensorFlow, DNN, RNN, LSTM}, + bdsk-url-1 = {https://doi.org/10.1145/3079856.3080246} +} +@article{Norrie2021TPUv2_3, + title = {The Design Process for Google's Training Chips: TPUv2 and TPUv3}, + author = {Norrie, Thomas and Patil, Nishant and Yoon, Doe Hyun and Kurian, George and Li, Sheng and Laudon, James and Young, Cliff and Jouppi, Norman and Patterson, David}, + year = 2021, + journal = {IEEE Micro}, + volume = 41, + number = 2, + pages = {56--63}, + doi = {10.1109/MM.2021.3058217}, + bdsk-url-1 = {https://doi.org/10.1109/MM.2021.3058217} +} @article{Northcutt_Athalye_Mueller_2021, - author = {Northcutt, Curtis G and Athalye, Anish and Mueller, Jonas}, - doi = {  https://doi.org/10.48550/arXiv.2103.14749 arXiv-issued DOI via DataCite}, - journal = {arXiv}, - month = {Mar}, - title = {Pervasive Label Errors in Test Sets Destabilize Machine Learning Benchmarks}, - year = {2021}, - Bdsk-Url-1 = { %20https://doi.org/10.48550/arXiv.2103.14749%20arXiv-issued%20DOI%20via%20DataCite}} - + title = {Pervasive Label Errors in Test Sets Destabilize Machine Learning Benchmarks}, + author = {Northcutt, Curtis G and Athalye, Anish and Mueller, Jonas}, + year = 2021, + month = {Mar}, + journal = {arXiv}, + doi = {  https://doi.org/10.48550/arXiv.2103.14749 arXiv-issued DOI via DataCite}, + bdsk-url-1 = { %20https://doi.org/10.48550/arXiv.2103.14749%20arXiv-issued%20DOI%20via%20DataCite} +} +@misc{obarBiggestLieInternet2018, + title = {The Biggest Lie on the Internet}, + author = {Obar, Jonathan A. and Oeldorf-Hirsch, Anne}, + year = 2018, + month = jun, + address = {Rochester, NY}, + doi = {10.2139/ssrn.2757465}, + url = {https://papers.ssrn.com/abstract=2757465}, + urldate = {2023-04-20}, + type = {{SSRN} {Scholarly} {Paper}}, + language = {en} +} +@article{obermeyer2019dissecting, + title = {Dissecting racial bias in an algorithm used to manage the health of populations}, + author = {Obermeyer, Ziad and Powers, Brian and Vogeli, Christine and Mullainathan, Sendhil}, + year = 2019, + journal = {Science}, + publisher = {American Association for the Advancement of Science}, + volume = 366, + number = 6464, + pages = {447--453} +} +@article{oecd22, + title = {Measuring the environmental impacts of artificial intelligence compute and applications}, + author = {OECD}, + year = 2022, + number = 341, + doi = {https://doi.org/https://doi.org/10.1787/7babf571-en}, + url = {https://www.oecd-ilibrary.org/content/paper/7babf571-en} +} +@article{olah2020zoom, + title = {Zoom in: An introduction to circuits}, + author = {Olah, Chris and Cammarata, Nick and Schubert, Ludwig and Goh, Gabriel and Petrov, Michael and Carter, Shan}, + year = 2020, + journal = {Distill}, + volume = 5, + number = 3, + pages = {e00024--001} +} +@article{oliynyk2023know, + title = {I Know What You Trained Last Summer: A Survey on Stealing Machine Learning Models and Defences}, + author = {Oliynyk, Daryna and Mayer, Rudolf and Rauber, Andreas}, + year = 2023, + month = {July}, + journal = {ACM Comput. Surv.}, + volume = 55, + number = {14s}, + date-added = {2023-11-22 16:18:21 -0500}, + date-modified = {2023-11-22 16:20:44 -0500}, + keywords = {model stealing, Machine learning, model extraction} +} @inproceedings{ooko2021tinyml, - author = {Ooko, Samson Otieno and Ogore, Marvin Muyonga and Nsenga, Jimmy and Zennaro, Marco}, - booktitle = {2021 IEEE Globecom Workshops (GC Wkshps)}, - organization = {IEEE}, - pages = {1--6}, - title = {TinyML in Africa: Opportunities and challenges}, - year = {2021}} - + title = {TinyML in Africa: Opportunities and challenges}, + author = {Ooko, Samson Otieno and Ogore, Marvin Muyonga and Nsenga, Jimmy and Zennaro, Marco}, + year = 2021, + booktitle = {2021 IEEE Globecom Workshops (GC Wkshps)}, + pages = {1--6}, + organization = {IEEE} +} +@article{oprea2022poisoning, + title = {Poisoning Attacks Against Machine Learning: Can Machine Learning Be Trustworthy?}, + author = {Oprea, Alina and Singhal, Anoop and Vassilev, Apostol}, + year = 2022, + journal = {Computer}, + publisher = {IEEE}, + volume = 55, + number = 11, + pages = {94--99} +} @misc{ou_low_2023, - abstract = {Deep neural networks have achieved great success in many data processing applications. However, the high computational complexity and storage cost makes deep learning hard to be used on resource-constrained devices, and it is not environmental-friendly with much power cost. In this paper, we focus on low-rank optimization for efficient deep learning techniques. In the space domain, deep neural networks are compressed by low rank approximation of the network parameters, which directly reduces the storage requirement with a smaller number of network parameters. In the time domain, the network parameters can be trained in a few subspaces, which enables efficient training for fast convergence. The model compression in the spatial domain is summarized into three categories as pre-train, pre-set, and compression-aware methods, respectively. With a series of integrable techniques discussed, such as sparse pruning, quantization, and entropy coding, we can ensemble them in an integration framework with lower computational complexity and storage. Besides of summary of recent technical advances, we have two findings for motivating future works: one is that the effective rank outperforms other sparse measures for network compression. The other is a spatial and temporal balance for tensorized neural networks.}, - author = {Ou, Xinwei and Chen, Zhangxin and Zhu, Ce and Liu, Yipeng}, - file = {arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/SPSZ2HR9/2303.html:text/html;Full Text PDF:/Users/jeffreyma/Zotero/storage/6TUEBTEX/Ou et al. - 2023 - Low Rank Optimization for Efficient Deep Learning.pdf:application/pdf}, - keywords = {Computer Science - Machine Learning}, - month = {Mar}, - note = {arXiv:2303.13635 [cs]}, - publisher = {arXiv}, - shorttitle = {Low {Rank} {Optimization} for {Efficient} {Deep} {Learning}}, - title = {Low {Rank} {Optimization} for {Efficient} {Deep} {Learning}: {Making} {A} {Balance} between {Compact} {Architecture} and {Fast} {Training}}, - url = {http://arxiv.org/abs/2303.13635}, - urldate = {2023-10-20}, - year = {2023}, - Bdsk-Url-1 = {http://arxiv.org/abs/2303.13635}} - + title = {Low {Rank} {Optimization} for {Efficient} {Deep} {Learning}: {Making} {A} {Balance} between {Compact} {Architecture} and {Fast} {Training}}, + shorttitle = {Low {Rank} {Optimization} for {Efficient} {Deep} {Learning}}, + author = {Ou, Xinwei and Chen, Zhangxin and Zhu, Ce and Liu, Yipeng}, + year = 2023, + month = {Mar}, + publisher = {arXiv}, + url = {http://arxiv.org/abs/2303.13635}, + urldate = {2023-10-20}, + note = {arXiv:2303.13635 [cs]}, + abstract = {Deep neural networks have achieved great success in many data processing applications. However, the high computational complexity and storage cost makes deep learning hard to be used on resource-constrained devices, and it is not environmental-friendly with much power cost. In this paper, we focus on low-rank optimization for efficient deep learning techniques. In the space domain, deep neural networks are compressed by low rank approximation of the network parameters, which directly reduces the storage requirement with a smaller number of network parameters. In the time domain, the network parameters can be trained in a few subspaces, which enables efficient training for fast convergence. The model compression in the spatial domain is summarized into three categories as pre-train, pre-set, and compression-aware methods, respectively. With a series of integrable techniques discussed, such as sparse pruning, quantization, and entropy coding, we can ensemble them in an integration framework with lower computational complexity and storage. Besides of summary of recent technical advances, we have two findings for motivating future works: one is that the effective rank outperforms other sparse measures for network compression. The other is a spatial and temporal balance for tensorized neural networks.}, + file = {arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/SPSZ2HR9/2303.html:text/html;Full Text PDF:/Users/jeffreyma/Zotero/storage/6TUEBTEX/Ou et al. - 2023 - Low Rank Optimization for Efficient Deep Learning.pdf:application/pdf}, + keywords = {Computer Science - Machine Learning}, + bdsk-url-1 = {http://arxiv.org/abs/2303.13635} +} @article{pan_survey_2010, - author = {Pan, Sinno Jialin and Yang, Qiang}, - doi = {10.1109/TKDE.2009.191}, - file = {Pan and Yang - 2010 - A Survey on Transfer Learning.pdf:/Users/alex/Zotero/storage/T3H8E5K8/Pan and Yang - 2010 - A Survey on Transfer Learning.pdf:application/pdf}, - issn = {1041-4347}, - journal = {IEEE Transactions on Knowledge and Data Engineering}, - language = {en}, - month = {Oct}, - number = {10}, - pages = {1345--1359}, - title = {A {Survey} on {Transfer} {Learning}}, - url = {http://ieeexplore.ieee.org/document/5288526/}, - urldate = {2023-10-25}, - volume = {22}, - year = {2010}, - Bdsk-Url-1 = {http://ieeexplore.ieee.org/document/5288526/}, - Bdsk-Url-2 = {https://doi.org/10.1109/TKDE.2009.191}} - + title = {A {Survey} on {Transfer} {Learning}}, + author = {Pan, Sinno Jialin and Yang, Qiang}, + year = 2010, + month = {Oct}, + journal = {IEEE Transactions on Knowledge and Data Engineering}, + publisher = {IEEE}, + volume = 22, + number = 10, + pages = {1345--1359}, + doi = {10.1109/TKDE.2009.191}, + issn = {1041-4347}, + url = {http://ieeexplore.ieee.org/document/5288526/}, + urldate = {2023-10-25}, + file = {Pan and Yang - 2010 - A Survey on Transfer Learning.pdf:/Users/alex/Zotero/storage/T3H8E5K8/Pan and Yang - 2010 - A Survey on Transfer Learning.pdf:application/pdf}, + language = {en}, + bdsk-url-1 = {http://ieeexplore.ieee.org/document/5288526/}, + bdsk-url-2 = {https://doi.org/10.1109/TKDE.2009.191} +} @article{pan2009survey, - author = {Pan, Sinno Jialin and Yang, Qiang}, - journal = {IEEE Transactions on knowledge and data engineering}, - number = {10}, - pages = {1345--1359}, - publisher = {IEEE}, - title = {A survey on transfer learning}, - volume = {22}, - year = {2009}} - + title = {A survey on transfer learning}, + author = {Pan, Sinno Jialin and Yang, Qiang}, + year = 2009, + journal = {IEEE Transactions on knowledge and data engineering}, + publisher = {IEEE}, + volume = 22, + number = 10, + pages = {1345--1359} +} +@inproceedings{papernot2017practical, + title = {Practical black-box attacks against machine learning}, + author = {Papernot, Nicolas and McDaniel, Patrick and Goodfellow, Ian and Jha, Somesh and Celik, Z Berkay and Swami, Ananthram}, + year = 2017, + booktitle = {Proceedings of the 2017 ACM on Asia conference on computer and communications security}, + pages = {506--519} +} @article{parisi_continual_2019, - author = {Parisi, German I. and Kemker, Ronald and Part, Jose L. and Kanan, Christopher and Wermter, Stefan}, - doi = {10.1016/j.neunet.2019.01.012}, - file = {Parisi et al. - 2019 - Continual lifelong learning with neural networks .pdf:/Users/alex/Zotero/storage/TCGHD5TW/Parisi et al. - 2019 - Continual lifelong learning with neural networks .pdf:application/pdf}, - issn = {08936080}, - journal = {Neural Networks}, - language = {en}, - month = {May}, - pages = {54--71}, - shorttitle = {Continual lifelong learning with neural networks}, - title = {Continual lifelong learning with neural networks: {A} review}, - url = {https://linkinghub.elsevier.com/retrieve/pii/S0893608019300231}, - urldate = {2023-10-26}, - volume = {113}, - year = {2019}, - Bdsk-Url-1 = {https://linkinghub.elsevier.com/retrieve/pii/S0893608019300231}, - Bdsk-Url-2 = {https://doi.org/10.1016/j.neunet.2019.01.012}} - + title = {Continual lifelong learning with neural networks: {A} review}, + shorttitle = {Continual lifelong learning with neural networks}, + author = {Parisi, German I. and Kemker, Ronald and Part, Jose L. and Kanan, Christopher and Wermter, Stefan}, + year = 2019, + month = {May}, + journal = {Neural Networks}, + volume = 113, + pages = {54--71}, + doi = {10.1016/j.neunet.2019.01.012}, + issn = {08936080}, + url = {https://linkinghub.elsevier.com/retrieve/pii/S0893608019300231}, + urldate = {2023-10-26}, + file = {Parisi et al. - 2019 - Continual lifelong learning with neural networks .pdf:/Users/alex/Zotero/storage/TCGHD5TW/Parisi et al. - 2019 - Continual lifelong learning with neural networks .pdf:application/pdf}, + language = {en}, + bdsk-url-1 = {https://linkinghub.elsevier.com/retrieve/pii/S0893608019300231}, + bdsk-url-2 = {https://doi.org/10.1016/j.neunet.2019.01.012} +} +@article{parrish2023adversarial, + title = {Adversarial Nibbler: A Data-Centric Challenge for Improving the Safety of Text-to-Image Models}, + author = {Alicia Parrish and Hannah Rose Kirk and Jessica Quaye and Charvi Rastogi and Max Bartolo and Oana Inel and Juan Ciro and Rafael Mosquera and Addison Howard and Will Cukierski and D. Sculley and Vijay Janapa Reddi and Lora Aroyo}, + year = 2023, + journal = {arXiv preprint arXiv:2305.14384}, + date-added = {2023-11-22 16:24:50 -0500}, + date-modified = {2023-11-22 16:26:30 -0500} +} @article{paszke2019pytorch, - author = {Paszke, Adam and Gross, Sam and Massa, Francisco and Lerer, Adam and Bradbury, James and Chanan, Gregory and Killeen, Trevor and Lin, Zeming and Gimelshein, Natalia and Antiga, Luca and others}, - journal = {Advances in neural information processing systems}, - title = {Pytorch: An imperative style, high-performance deep learning library}, - volume = {32}, - year = {2019}} - + title = {Pytorch: An imperative style, high-performance deep learning library}, + author = {Paszke, Adam and Gross, Sam and Massa, Francisco and Lerer, Adam and Bradbury, James and Chanan, Gregory and Killeen, Trevor and Lin, Zeming and Gimelshein, Natalia and Antiga, Luca and others}, + year = 2019, + journal = {Advances in neural information processing systems}, + volume = 32 +} +@book{patterson2016computer, + title = {Computer organization and design ARM edition: the hardware software interface}, + author = {Patterson, David A and Hennessy, John L}, + year = 2016, + publisher = {Morgan kaufmann} +} +@article{patterson2022carbon, + title = {The carbon footprint of machine learning training will plateau, then shrink}, + author = {Patterson, David and Gonzalez, Joseph and H{\"o}lzle, Urs and Le, Quoc and Liang, Chen and Munguia, Lluis-Miquel and Rothchild, Daniel and So, David R and Texier, Maud and Dean, Jeff}, + year = 2022, + journal = {Computer}, + publisher = {IEEE}, + volume = 55, + number = 7, + pages = {18--28} +} @misc{Perrigo_2023, - author = {Perrigo, Billy}, - journal = {Time}, - month = {Jan}, - publisher = {Time}, - title = {OpenAI used Kenyan workers on less than $2 per hour: Exclusive}, - url = {https://time.com/6247678/openai-chatgpt-kenya-workers/}, - year = {2023}, - Bdsk-Url-1 = {https://time.com/6247678/openai-chatgpt-kenya-workers/}} - + title = {OpenAI used Kenyan workers on less than $2 per hour: Exclusive}, + author = {Perrigo, Billy}, + year = 2023, + month = {Jan}, + journal = {Time}, + publisher = {Time}, + url = {https://time.com/6247678/openai-chatgpt-kenya-workers/}, + bdsk-url-1 = {https://time.com/6247678/openai-chatgpt-kenya-workers/} +} +@article{peters2018designing, + title = {Designing for motivation, engagement and wellbeing in digital experience}, + author = {Peters, Dorian and Calvo, Rafael A and Ryan, Richard M}, + year = 2018, + journal = {Frontiers in psychology}, + publisher = {Frontiers}, + pages = 797 +} +@article{phillips2020four, + title = {Four principles of explainable artificial intelligence}, + author = {Phillips, P Jonathon and Hahn, Carina A and Fontana, Peter C and Broniatowski, David A and Przybocki, Mark A}, + year = 2020, + journal = {Gaithersburg, Maryland}, + volume = 18 +} +@article{plasma, + title = {Noninvasive assessment of dofetilide plasma concentration using a deep learning (neural network) analysis of the surface electrocardiogram: A proof of concept study}, + author = {Attia, Zachi and Sugrue, Alan and Asirvatham, Samuel and Ackerman, Michael and Kapa, Suraj and Friedman, Paul and Noseworthy, Peter}, + year = 2018, + month = {08}, + journal = {PLOS ONE}, + volume = 13, + pages = {e0201059}, + doi = {10.1371/journal.pone.0201059}, + bdsk-url-1 = {https://doi.org/10.1371/journal.pone.0201059} +} +@article{poff2002aquatic, + title = {Aquatic ecosystems \& Global climate change}, + author = {LeRoy Poff, N and Brinson, MM and Day, JW}, + year = 2002, + journal = {Pew Center on Global Climate Change} +} @inproceedings{Prakash_2023, - author = {Shvetank Prakash and Tim Callahan and Joseph Bushagour and Colby Banbury and Alan V. Green and Pete Warden and Tim Ansell and Vijay Janapa Reddi}, - booktitle = {2023 {IEEE} International Symposium on Performance Analysis of Systems and Software ({ISPASS})}, - doi = {10.1109/ispass57527.2023.00024}, - month = {apr}, - publisher = {{IEEE}}, - title = {{CFU} Playground: Full-Stack Open-Source Framework for Tiny Machine Learning ({TinyML}) Acceleration on {FPGAs}}, - url = {https://doi.org/10.1109%2Fispass57527.2023.00024}, - year = {2023}, - Bdsk-Url-1 = {https://doi.org/10.1109%2Fispass57527.2023.00024}, - Bdsk-Url-2 = {https://doi.org/10.1109/ispass57527.2023.00024}} - + title = {{CFU} Playground: Full-Stack Open-Source Framework for Tiny Machine Learning ({TinyML}) Acceleration on {FPGAs}}, + shorttitle = {{CFU} {Playground}}, + author = {Shvetank Prakash and Tim Callahan and Joseph Bushagour and Colby Banbury and Alan V. Green and Pete Warden and Tim Ansell and Vijay Janapa Reddi}, + year = 2023, + month = {apr}, + booktitle = {2023 {IEEE} International Symposium on Performance Analysis of Systems and Software ({ISPASS})}, + publisher = {{IEEE}}, + pages = {157--167}, + doi = {10.1109/ispass57527.2023.00024}, + url = {https://doi.org/10.1109%2Fispass57527.2023.00024}, + urldate = {2023-10-25}, + note = {arXiv:2201.01863 [cs]}, + bdsk-url-1 = {https://doi.org/10.1109%2Fispass57527.2023.00024}, + bdsk-url-2 = {https://doi.org/10.1109/ispass57527.2023.00024}, + file = {Prakash et al. - 2023 - CFU Playground Full-Stack Open-Source Framework f.pdf:/Users/alex/Zotero/storage/BZNRIDTL/Prakash et al. - 2023 - CFU Playground Full-Stack Open-Source Framework f.pdf:application/pdf}, + keywords = {Computer Science - Machine Learning, Computer Science - Distributed, Parallel, and Cluster Computing, Computer Science - Hardware Architecture}, + language = {en} +} @inproceedings{prakash_cfu_2023, - author = {Prakash, Shvetank and Callahan, Tim and Bushagour, Joseph and Banbury, Colby and Green, Alan V. and Warden, Pete and Ansell, Tim and Reddi, Vijay Janapa}, - booktitle = {2023 {IEEE} {International} {Symposium} on {Performance} {Analysis} of {Systems} and {Software} ({ISPASS})}, - doi = {10.1109/ISPASS57527.2023.00024}, - file = {Prakash et al. - 2023 - CFU Playground Full-Stack Open-Source Framework f.pdf:/Users/alex/Zotero/storage/BZNRIDTL/Prakash et al. - 2023 - CFU Playground Full-Stack Open-Source Framework f.pdf:application/pdf}, - keywords = {Computer Science - Machine Learning, Computer Science - Distributed, Parallel, and Cluster Computing, Computer Science - Hardware Architecture}, - language = {en}, - month = {Apr}, - note = {arXiv:2201.01863 [cs]}, - pages = {157--167}, - shorttitle = {{CFU} {Playground}}, - title = {{CFU} {Playground}: {Full}-{Stack} {Open}-{Source} {Framework} for {Tiny} {Machine} {Learning} ({tinyML}) {Acceleration} on {FPGAs}}, - url = {http://arxiv.org/abs/2201.01863}, - urldate = {2023-10-25}, - year = {2023}, - Bdsk-Url-1 = {http://arxiv.org/abs/2201.01863}, - Bdsk-Url-2 = {https://doi.org/10.1109/ISPASS57527.2023.00024}} - + title = {{CFU} {Playground}: {Full}-{Stack} {Open}-{Source} {Framework} for {Tiny} {Machine} {Learning} ({tinyML}) {Acceleration} on {FPGAs}}, + shorttitle = {{CFU} {Playground}}, + author = {Prakash, Shvetank and Callahan, Tim and Bushagour, Joseph and Banbury, Colby and Green, Alan V. and Warden, Pete and Ansell, Tim and Reddi, Vijay Janapa}, + year = 2023, + month = {Apr}, + booktitle = {2023 {IEEE} {International} {Symposium} on {Performance} {Analysis} of {Systems} and {Software} ({ISPASS})}, + pages = {157--167}, + doi = {10.1109/ISPASS57527.2023.00024}, + url = {http://arxiv.org/abs/2201.01863}, + urldate = {2023-10-25}, + note = {arXiv:2201.01863 [cs]}, + language = {en}, + keywords = {Computer Science - Machine Learning, Computer Science - Distributed, Parallel, and Cluster Computing, Computer Science - Hardware Architecture}, + file = {Prakash et al. - 2023 - CFU Playground Full-Stack Open-Source Framework f.pdf:/Users/alex/Zotero/storage/BZNRIDTL/Prakash et al. - 2023 - CFU Playground Full-Stack Open-Source Framework f.pdf:application/pdf}, + bdsk-url-1 = {http://arxiv.org/abs/2201.01863}, + bdsk-url-2 = {https://doi.org/10.1109/ISPASS57527.2023.00024} +} +@misc{prakash2023tinyml, + title = {Is TinyML Sustainable? Assessing the Environmental Impacts of Machine Learning on Microcontrollers}, + author = {Shvetank Prakash and Matthew Stewart and Colby Banbury and Mark Mazumder and Pete Warden and Brian Plancher and Vijay Janapa Reddi}, + year = 2023, + journal = {arXiv preprint arXiv:2301.11899}, + eprint = {2301.11899}, + archiveprefix = {arXiv}, + primaryclass = {cs.LG} +} +@misc{prakashTinyMLSustainableAssessing2023, + title = {Is {TinyML} {Sustainable}?}, + author = {Prakash, Shvetank and Stewart, Matthew and Banbury, Colby and Mazumder, Mark and Warden, Pete and Plancher, Brian and Reddi, Vijay Janapa}, + year = 2023, + month = may, + publisher = {arXiv}, + url = {http://arxiv.org/abs/2301.11899}, + urldate = {2023-11-12}, + note = {arXiv:2301.11899 [cs]}, + annote = {Comment: To appear in Communications of the ACM (CACM) in 2023} +} @article{preparednesspublic, - author = {Preparedness, Emergency}, - title = {Public Health Law}} - + title = {Public Health Law}, + author = {Preparedness, Emergency} +} @article{Pushkarna_Zaldivar_Kjartansson_2022, - author = {Pushkarna, Mahima and Zaldivar, Andrew and Kjartansson, Oddur}, - doi = {10.1145/3531146.3533231}, - journal = {2022 ACM Conference on Fairness, Accountability, and Transparency}, - title = {Data cards: Purposeful and transparent dataset documentation for responsible ai}, - year = {2022}, - Bdsk-Url-1 = {https://doi.org/10.1145/3531146.3533231}} - + title = {Data cards: Purposeful and transparent dataset documentation for responsible ai}, + author = {Pushkarna, Mahima and Zaldivar, Andrew and Kjartansson, Oddur}, + year = 2022, + journal = {2022 ACM Conference on Fairness, Accountability, and Transparency}, + doi = {10.1145/3531146.3533231}, + bdsk-url-1 = {https://doi.org/10.1145/3531146.3533231} +} +@article{putnam_reconfigurable_2014, + title = {A reconfigurable fabric for accelerating large-scale datacenter services}, + author = {Putnam, Andrew and Caulfield, Adrian M. and Chung, Eric S. and Chiou, Derek and Constantinides, Kypros and Demme, John and Esmaeilzadeh, Hadi and Fowers, Jeremy and Gopal, Gopi Prashanth and Gray, Jan and Haselman, Michael and Hauck, Scott and Heil, Stephen and Hormati, Amir and Kim, Joo-Young and Lanka, Sitaram and Larus, James and Peterson, Eric and Pope, Simon and Smith, Aaron and Thong, Jason and Xiao, Phillip Yi and Burger, Doug}, + year = 2014, + month = oct, + journal = {ACM SIGARCH Computer Architecture News}, + volume = 42, + number = 3, + pages = {13--24}, + doi = {10.1145/2678373.2665678}, + issn = {0163-5964}, + url = {https://dl.acm.org/doi/10.1145/2678373.2665678}, + urldate = {2023-11-07}, + abstract = {Datacenter workloads demand high computational capabilities, flexibility, power efficiency, and low cost. It is challenging to improve all of these factors simultaneously. To advance datacenter capabilities beyond what commodity server designs can provide, we have designed and built a composable, reconfigurablefabric to accelerate portions of large-scale software services. Each instantiation of the fabric consists of a 6x8 2-D torus of high-end Stratix V FPGAs embedded into a half-rack of 48 machines. One FPGA is placed into each server, accessible through PCIe, and wired directly to other FPGAs with pairs of 10 Gb SAS cables In this paper, we describe a medium-scale deployment of this fabric on a bed of 1,632 servers, and measure its efficacy in accelerating the Bing web search engine. We describe the requirements and architecture of the system, detail the critical engineering challenges and solutions needed to make the system robust in the presence of failures, and measure the performance, power, and resilience of the system when ranking candidate documents. Under high load, the largescale reconfigurable fabric improves the ranking throughput of each server by a factor of 95\% for a fixed latency distribution--- or, while maintaining equivalent throughput, reduces the tail latency by 29\%}, + language = {en}, + bdsk-url-1 = {https://dl.acm.org/doi/10.1145/2678373.2665678}, + bdsk-url-2 = {https://doi.org/10.1145/2678373.2665678} +} @article{qi_efficient_2021, - abstract = {Nowadays, deep neural networks (DNNs) have been rapidly deployed to realize a number of functionalities like sensing, imaging, classification, recognition, etc. However, the computational-intensive requirement of DNNs makes it difficult to be applicable for resource-limited Internet of Things (IoT) devices. In this paper, we propose a novel pruning-based paradigm that aims to reduce the computational cost of DNNs, by uncovering a more compact structure and learning the effective weights therein, on the basis of not compromising the expressive capability of DNNs. In particular, our algorithm can achieve efficient end-to-end training that transfers a redundant neural network to a compact one with a specifically targeted compression rate directly. We comprehensively evaluate our approach on various representative benchmark datasets and compared with typical advanced convolutional neural network (CNN) architectures. The experimental results verify the superior performance and robust effectiveness of our scheme. For example, when pruning VGG on CIFAR-10, our proposed scheme is able to significantly reduce its FLOPs (floating-point operations) and number of parameters with a proportion of 76.2\% and 94.1\%, respectively, while still maintaining a satisfactory accuracy. To sum up, our scheme could facilitate the integration of DNNs into the common machine-learning-based IoT framework and establish distributed training of neural networks in both cloud and edge.}, - author = {Qi, Chen and Shen, Shibo and Li, Rongpeng and Zhifeng, Zhao and Liu, Qing and Liang, Jing and Zhang, Honggang}, - doi = {10.1186/s13634-021-00744-4}, - file = {Full Text PDF:/Users/jeffreyma/Zotero/storage/AGWCC5VS/Qi et al. - 2021 - An efficient pruning scheme of deep neural network.pdf:application/pdf}, - journal = {EURASIP Journal on Advances in Signal Processing}, - month = {Jun}, - title = {An efficient pruning scheme of deep neural networks for {Internet} of {Things} applications}, - volume = 2021, - year = {2021}, - Bdsk-Url-1 = {https://doi.org/10.1186/s13634-021-00744-4}} - + title = {An efficient pruning scheme of deep neural networks for {Internet} of {Things} applications}, + author = {Qi, Chen and Shen, Shibo and Li, Rongpeng and Zhifeng, Zhao and Liu, Qing and Liang, Jing and Zhang, Honggang}, + year = 2021, + month = {Jun}, + journal = {EURASIP Journal on Advances in Signal Processing}, + volume = 2021, + doi = {10.1186/s13634-021-00744-4}, + abstract = {Nowadays, deep neural networks (DNNs) have been rapidly deployed to realize a number of functionalities like sensing, imaging, classification, recognition, etc. However, the computational-intensive requirement of DNNs makes it difficult to be applicable for resource-limited Internet of Things (IoT) devices. In this paper, we propose a novel pruning-based paradigm that aims to reduce the computational cost of DNNs, by uncovering a more compact structure and learning the effective weights therein, on the basis of not compromising the expressive capability of DNNs. In particular, our algorithm can achieve efficient end-to-end training that transfers a redundant neural network to a compact one with a specifically targeted compression rate directly. We comprehensively evaluate our approach on various representative benchmark datasets and compared with typical advanced convolutional neural network (CNN) architectures. The experimental results verify the superior performance and robust effectiveness of our scheme. For example, when pruning VGG on CIFAR-10, our proposed scheme is able to significantly reduce its FLOPs (floating-point operations) and number of parameters with a proportion of 76.2\% and 94.1\%, respectively, while still maintaining a satisfactory accuracy. To sum up, our scheme could facilitate the integration of DNNs into the common machine-learning-based IoT framework and establish distributed training of neural networks in both cloud and edge.}, + file = {Full Text PDF:/Users/jeffreyma/Zotero/storage/AGWCC5VS/Qi et al. - 2021 - An efficient pruning scheme of deep neural network.pdf:application/pdf}, + bdsk-url-1 = {https://doi.org/10.1186/s13634-021-00744-4} +} @misc{quantdeep, - author = {Krishnamoorthi}, - doi = {10.48550/arXiv.1806.08342}, - month = jun, - publisher = {arXiv}, - title = {Quantizing deep convolutional networks for efficient inference: A whitepaper}, - url = {https://arxiv.org/abs/1806.08342}, - urldate = {2018-06-21}, - year = 2018, - Bdsk-Url-1 = {https://arxiv.org/abs/1806.08342}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.1806.08342}} - + title = {Quantizing deep convolutional networks for efficient inference: A whitepaper}, + author = {Krishnamoorthi}, + year = 2018, + month = jun, + publisher = {arXiv}, + doi = {10.48550/arXiv.1806.08342}, + url = {https://arxiv.org/abs/1806.08342}, + urldate = {2018-06-21}, + bdsk-url-1 = {https://arxiv.org/abs/1806.08342}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.1806.08342} +} +@inproceedings{raina_large-scale_2009, + title = {Large-scale deep unsupervised learning using graphics processors}, + author = {Raina, Rajat and Madhavan, Anand and Ng, Andrew Y.}, + year = 2009, + month = jun, + booktitle = {Proceedings of the 26th {Annual} {International} {Conference} on {Machine} {Learning}}, + publisher = {ACM}, + address = {Montreal Quebec Canada}, + pages = {873--880}, + doi = {10.1145/1553374.1553486}, + isbn = {978-1-60558-516-1}, + url = {https://dl.acm.org/doi/10.1145/1553374.1553486}, + urldate = {2023-11-07}, + language = {en}, + bdsk-url-1 = {https://dl.acm.org/doi/10.1145/1553374.1553486}, + bdsk-url-2 = {https://doi.org/10.1145/1553374.1553486} +} +@inproceedings{ramaswamy2023overlooked, + title = {Overlooked Factors in Concept-Based Explanations: Dataset Choice, Concept Learnability, and Human Capability}, + author = {Ramaswamy, Vikram V and Kim, Sunnie SY and Fong, Ruth and Russakovsky, Olga}, + year = 2023, + booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition}, + pages = {10932--10941} +} +@article{ramaswamy2023ufo, + title = {UFO: A unified method for controlling Understandability and Faithfulness Objectives in concept-based explanations for CNNs}, + author = {Ramaswamy, Vikram V and Kim, Sunnie SY and Fong, Ruth and Russakovsky, Olga}, + year = 2023, + journal = {arXiv preprint arXiv:2303.15632} +} @article{ramcharan2017deep, - author = {Ramcharan, Amanda and Baranowski, Kelsee and McCloskey, Peter and Ahmed, Babuali and Legg, James and Hughes, David P}, - journal = {Frontiers in plant science}, - pages = 1852, - publisher = {Frontiers Media SA}, - title = {Deep learning for image-based cassava disease detection}, - volume = 8, - year = 2017} - + title = {Deep learning for image-based cassava disease detection}, + author = {Ramcharan, Amanda and Baranowski, Kelsee and McCloskey, Peter and Ahmed, Babuali and Legg, James and Hughes, David P}, + year = 2017, + journal = {Frontiers in plant science}, + publisher = {Frontiers Media SA}, + volume = 8, + pages = 1852 +} +@inproceedings{ramesh2021zero, + title = {Zero-shot text-to-image generation}, + author = {Ramesh, Aditya and Pavlov, Mikhail and Goh, Gabriel and Gray, Scott and Voss, Chelsea and Radford, Alec and Chen, Mark and Sutskever, Ilya}, + year = 2021, + booktitle = {International Conference on Machine Learning}, + pages = {8821--8831}, + organization = {PMLR} +} +@article{Ranganathan2011-dc, + title = {From microprocessors to nanostores: Rethinking data-centric systems}, + author = {Ranganathan, Parthasarathy}, + year = 2011, + month = jan, + journal = {Computer (Long Beach Calif.)}, + publisher = {Institute of Electrical and Electronics Engineers (IEEE)}, + volume = 44, + number = 1, + pages = {39--48} +} @misc{Rao_2021, - author = {Rao, Ravi}, - journal = {www.wevolver.com}, - month = {Dec}, - url = {https://www.wevolver.com/article/tinyml-unlocks-new-possibilities-for-sustainable-development-technologies}, - year = 2021, - Bdsk-Url-1 = {https://www.wevolver.com/article/tinyml-unlocks-new-possibilities-for-sustainable-development-technologies}} - + author = {Rao, Ravi}, + year = 2021, + month = {Dec}, + journal = {www.wevolver.com}, + url = {https://www.wevolver.com/article/tinyml-unlocks-new-possibilities-for-sustainable-development-technologies}, + bdsk-url-1 = {https://www.wevolver.com/article/tinyml-unlocks-new-possibilities-for-sustainable-development-technologies} +} +@inproceedings{Rashmi2018Secure, + title = {Secure boot of Embedded Applications - A Review}, + author = {R.V. Rashmi and A. Karthikeyan}, + year = 2018, + booktitle = {2018 Second International Conference on Electronics, Communication and Aerospace Technology (ICECA)}, + pages = {291--298}, + date-added = {2023-11-22 17:50:16 -0500}, + date-modified = {2023-11-22 17:51:39 -0500} +} +@article{Ratner_Hancock_Dunnmon_Goldman_Ré_2018, + title = {Snorkel metal: Weak supervision for multi-task learning.}, + author = {Ratner, Alex and Hancock, Braden and Dunnmon, Jared and Goldman, Roger and R\'{e}, Christopher}, + year = 2018, + journal = {Proceedings of the Second Workshop on Data Management for End-To-End Machine Learning}, + doi = {10.1145/3209889.3209898} +} @inproceedings{reddi2020mlperf, - author = {Reddi, Vijay Janapa and Cheng, Christine and Kanter, David and Mattson, Peter and Schmuelling, Guenther and Wu, Carole-Jean and Anderson, Brian and Breughe, Maximilien and Charlebois, Mark and Chou, William and others}, - booktitle = {2020 ACM/IEEE 47th Annual International Symposium on Computer Architecture (ISCA)}, - organization = {IEEE}, - pages = {446--459}, - title = {Mlperf inference benchmark}, - year = 2020} - + title = {Mlperf inference benchmark}, + author = {Reddi, Vijay Janapa and Cheng, Christine and Kanter, David and Mattson, Peter and Schmuelling, Guenther and Wu, Carole-Jean and Anderson, Brian and Breughe, Maximilien and Charlebois, Mark and Chou, William and others}, + year = 2020, + booktitle = {2020 ACM/IEEE 47th Annual International Symposium on Computer Architecture (ISCA)}, + pages = {446--459}, + organization = {IEEE} +} @inproceedings{ribeiro2016should, - author = {Ribeiro, Marco Tulio and Singh, Sameer and Guestrin, Carlos}, - booktitle = {Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining}, - pages = {1135--1144}, - title = {" Why should i trust you?" Explaining the predictions of any classifier}, - year = 2016} - + title = {" Why should i trust you?" Explaining the predictions of any classifier}, + author = {Ribeiro, Marco Tulio and Singh, Sameer and Guestrin, Carlos}, + year = 2016, + booktitle = {Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining}, + pages = {1135--1144} +} +@misc{rmsprop, + title = {Overview of Minibatch Gradient Descent}, + author = {Geoffrey Hinton}, + year = 2017, + institution = {University of Toronto}, + howpublished = {University Lecture} +} +@conference{Rombach22cvpr, + title = {High-Resolution Image Synthesis with Latent Diffusion Models}, + author = {Robin Rombach and Andreas Blattmann and Dominik Lorenz and Patrick Esser and Björn Ommer}, + year = 2022, + booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)}, + url = {https://github.com/CompVis/latent-diffusionhttps://arxiv.org/abs/2112.10752} +} @book{rosenblatt1957perceptron, - author = {Rosenblatt, Frank}, - publisher = {Cornell Aeronautical Laboratory}, - title = {The perceptron, a perceiving and recognizing automaton Project Para}, - year = 1957} - + title = {The perceptron, a perceiving and recognizing automaton Project Para}, + author = {Rosenblatt, Frank}, + year = 1957, + publisher = {Cornell Aeronautical Laboratory} +} +@article{roskies2002neuroethics, + title = {Neuroethics for the new millenium}, + author = {Roskies, Adina}, + year = 2002, + journal = {Neuron}, + publisher = {Elsevier}, + volume = 35, + number = 1, + pages = {21--23} +} +@inproceedings{ross2018improving, + title = {Improving the adversarial robustness and interpretability of deep neural networks by regularizing their input gradients}, + author = {Ross, Andrew and Doshi-Velez, Finale}, + year = 2018, + booktitle = {Proceedings of the AAAI conference on artificial intelligence}, + volume = 32, + number = 1 +} @inproceedings{rouhani2017tinydl, - author = {Rouhani, Bita and Mirhoseini, Azalia and Koushanfar, Farinaz}, - doi = {10.1109/ISCAS.2017.8050343}, - month = {05}, - pages = {1--4}, - title = {TinyDL: Just-in-time deep learning solution for constrained embedded systems}, - year = 2017, - Bdsk-Url-1 = {https://doi.org/10.1109/ISCAS.2017.8050343}} - + title = {TinyDL: Just-in-time deep learning solution for constrained embedded systems}, + author = {Rouhani, Bita and Mirhoseini, Azalia and Koushanfar, Farinaz}, + year = 2017, + month = {05}, + pages = {1--4}, + doi = {10.1109/ISCAS.2017.8050343}, + bdsk-url-1 = {https://doi.org/10.1109/ISCAS.2017.8050343} +} +@article{ruder2016overview, + title = {An overview of gradient descent optimization algorithms}, + author = {Ruder, Sebastian}, + year = 2016, + journal = {arXiv preprint arXiv:1609.04747} +} +@article{rudin2019stop, + title = {Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead}, + author = {Rudin, Cynthia}, + year = 2019, + journal = {Nature machine intelligence}, + publisher = {Nature Publishing Group UK London}, + volume = 1, + number = 5, + pages = {206--215} +} @article{rumelhart1986learning, - author = {Rumelhart, David E and Hinton, Geoffrey E and Williams, Ronald J}, - journal = {nature}, - number = 6088, - pages = {533--536}, - publisher = {Nature Publishing Group UK London}, - title = {Learning representations by back-propagating errors}, - volume = 323, - year = 1986} - + title = {Learning representations by back-propagating errors}, + author = {Rumelhart, David E and Hinton, Geoffrey E and Williams, Ronald J}, + year = 1986, + journal = {nature}, + publisher = {Nature Publishing Group UK London}, + volume = 323, + number = 6088, + pages = {533--536} +} +@article{russell2021human, + title = {Human-compatible artificial intelligence}, + author = {Russell, Stuart}, + year = 2021, + journal = {Human-like machine intelligence}, + publisher = {Oxford University Press Oxford}, + pages = {3--23} +} @article{ruvolo_ella_nodate, - author = {Ruvolo, Paul and Eaton, Eric}, - file = {Ruvolo and Eaton - ELLA An Efficient Lifelong Learning Algorithm.pdf:/Users/alex/Zotero/storage/QA5G29GL/Ruvolo and Eaton - ELLA An Efficient Lifelong Learning Algorithm.pdf:application/pdf}, - language = {en}, - title = {{ELLA}: {An} {Efficient} {Lifelong} {Learning} {Algorithm}}} - + title = {{ELLA}: {An} {Efficient} {Lifelong} {Learning} {Algorithm}}, + author = {Ruvolo, Paul and Eaton, Eric}, + file = {Ruvolo and Eaton - ELLA An Efficient Lifelong Learning Algorithm.pdf:/Users/alex/Zotero/storage/QA5G29GL/Ruvolo and Eaton - ELLA An Efficient Lifelong Learning Algorithm.pdf:application/pdf}, + language = {en} +} +@article{ryan2000self, + title = {Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being.}, + author = {Ryan, Richard M and Deci, Edward L}, + year = 2000, + journal = {American psychologist}, + publisher = {American Psychological Association}, + volume = 55, + number = 1, + pages = 68 +} +@article{samajdar2018scale, + title = {Scale-sim: Systolic cnn accelerator simulator}, + author = {Samajdar, Ananda and Zhu, Yuhao and Whatmough, Paul and Mattina, Matthew and Krishna, Tushar}, + year = 2018, + journal = {arXiv preprint arXiv:1811.02883} +} @misc{ScaleAI, - journal = {ScaleAI}, - url = {https://scale.com/data-engine}, - Bdsk-Url-1 = {https://scale.com/data-engine}} - + journal = {ScaleAI}, + url = {https://scale.com/data-engine}, + bdsk-url-1 = {https://scale.com/data-engine} +} +@article{scaling_laws_NLM, + title = {Scaling Laws for Neural Language Models}, + author = {Jared Kaplan and Sam McCandlish and Tom Henighan and Tom B. Brown and Benjamin Chess and Rewon Child and Scott Gray and Alec Radford and Jeffrey Wu and Dario Amodei}, + year = 2020, + journal = {CoRR}, + volume = {abs/2001.08361}, + url = {https://arxiv.org/abs/2001.08361}, + eprinttype = {arXiv}, + eprint = {2001.08361}, + timestamp = {Wed, 03 Jun 2020 10:55:13 +0200}, + biburl = {https://dblp.org/rec/journals/corr/abs-2001-08361.bib}, + bibsource = {dblp computer science bibliography, https://dblp.org} +} +@article{schafer2023notorious, + title = {The Notorious GPT: science communication in the age of artificial intelligence}, + author = {Sch{\"a}fer, Mike S}, + year = 2023, + journal = {Journal of Science Communication}, + publisher = {SISSA Medialab srl}, + volume = 22, + number = 2, + pages = {Y02} +} +@article{schuman2022, + title = {Opportunities for neuromorphic computing algorithms and applications}, + author = {Schuman, Catherine D and Kulkarni, Shruti R and Parsa, Maryam and Mitchell, J Parker and Date, Prasanna and Kay, Bill}, + year = 2022, + journal = {Nature Computational Science}, + publisher = {Nature Publishing Group US New York}, + volume = 2, + number = 1, + pages = {10--19} +} +@article{schwartz2020green, + title = {Green ai}, + author = {Schwartz, Roy and Dodge, Jesse and Smith, Noah A and Etzioni, Oren}, + year = 2020, + journal = {Communications of the ACM}, + publisher = {ACM New York, NY, USA}, + volume = 63, + number = 12, + pages = {54--63} +} +@inproceedings{schwartz2021deployment, + title = {Deployment of Embedded Edge-AI for Wildlife Monitoring in Remote Regions}, + author = {Schwartz, Daniel and Selman, Jonathan Michael Gomes and Wrege, Peter and Paepcke, Andreas}, + year = 2021, + booktitle = {2021 20th IEEE International Conference on Machine Learning and Applications (ICMLA)}, + pages = {1035--1042}, + organization = {IEEE} +} +@inproceedings{schwartzDeploymentEmbeddedEdgeAI2021, + title = {Deployment of Embedded Edge-AI for Wildlife Monitoring in Remote Regions}, + author = {Schwartz, Daniel and Selman, Jonathan Michael Gomes and Wrege, Peter and Paepcke, Andreas}, + year = 2021, + booktitle = {2021 20th IEEE International Conference on Machine Learning and Applications (ICMLA)}, + pages = {1035--1042}, + organization = {IEEE} +} @inproceedings{schwarzschild2021just, - author = {Schwarzschild, Avi and Goldblum, Micah and Gupta, Arjun and Dickerson, John P and Goldstein, Tom}, - booktitle = {International Conference on Machine Learning}, - organization = {PMLR}, - pages = {9389--9398}, - title = {Just how toxic is data poisoning? a unified benchmark for backdoor and data poisoning attacks}, - year = 2021} - + title = {Just how toxic is data poisoning? a unified benchmark for backdoor and data poisoning attacks}, + author = {Schwarzschild, Avi and Goldblum, Micah and Gupta, Arjun and Dickerson, John P and Goldstein, Tom}, + year = 2021, + booktitle = {International Conference on Machine Learning}, + pages = {9389--9398}, + organization = {PMLR} +} +@inproceedings{sculley2015hidden, + title = {"Everyone wants to do the model work, not the data work": Data Cascades in High-Stakes AI}, + author = {Nithya Sambasivan and Shivani Kapania and Hannah Highfill and Diana Akrong and Praveen Kumar Paritosh and Lora Mois Aroyo}, + year = 2021 +} @misc{see_compression_2016, - abstract = {Neural Machine Translation (NMT), like many other deep learning domains, typically suffers from over-parameterization, resulting in large storage sizes. This paper examines three simple magnitude-based pruning schemes to compress NMT models, namely class-blind, class-uniform, and class-distribution, which differ in terms of how pruning thresholds are computed for the different classes of weights in the NMT architecture. We demonstrate the efficacy of weight pruning as a compression technique for a state-of-the-art NMT system. We show that an NMT model with over 200 million parameters can be pruned by 40\% with very little performance loss as measured on the WMT'14 English-German translation task. This sheds light on the distribution of redundancy in the NMT architecture. Our main result is that with retraining, we can recover and even surpass the original performance with an 80\%-pruned model.}, - author = {See, Abigail and Luong, Minh-Thang and Manning, Christopher D.}, - doi = {10.48550/arXiv.1606.09274}, - file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/2CJ4TSNR/See et al. - 2016 - Compression of Neural Machine Translation Models v.pdf:application/pdf}, - keywords = {Computer Science - Artificial Intelligence, Computer Science - Computation and Language, Computer Science - Neural and Evolutionary Computing}, - month = jun, - note = {arXiv:1606.09274 [cs]}, - publisher = {arXiv}, - title = {Compression of {Neural} {Machine} {Translation} {Models} via {Pruning}}, - url = {http://arxiv.org/abs/1606.09274}, - urldate = {2023-10-20}, - year = 2016, - Bdsk-Url-1 = {http://arxiv.org/abs/1606.09274}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.1606.09274}} - + title = {Compression of {Neural} {Machine} {Translation} {Models} via {Pruning}}, + author = {See, Abigail and Luong, Minh-Thang and Manning, Christopher D.}, + year = 2016, + month = jun, + publisher = {arXiv}, + doi = {10.48550/arXiv.1606.09274}, + url = {http://arxiv.org/abs/1606.09274}, + urldate = {2023-10-20}, + note = {arXiv:1606.09274 [cs]}, + abstract = {Neural Machine Translation (NMT), like many other deep learning domains, typically suffers from over-parameterization, resulting in large storage sizes. This paper examines three simple magnitude-based pruning schemes to compress NMT models, namely class-blind, class-uniform, and class-distribution, which differ in terms of how pruning thresholds are computed for the different classes of weights in the NMT architecture. We demonstrate the efficacy of weight pruning as a compression technique for a state-of-the-art NMT system. We show that an NMT model with over 200 million parameters can be pruned by 40\% with very little performance loss as measured on the WMT'14 English-German translation task. This sheds light on the distribution of redundancy in the NMT architecture. Our main result is that with retraining, we can recover and even surpass the original performance with an 80\%-pruned model.}, + file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/2CJ4TSNR/See et al. - 2016 - Compression of Neural Machine Translation Models v.pdf:application/pdf}, + keywords = {Computer Science - Artificial Intelligence, Computer Science - Computation and Language, Computer Science - Neural and Evolutionary Computing}, + bdsk-url-1 = {http://arxiv.org/abs/1606.09274}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.1606.09274} +} +@misc{segal1999opengl, + title = {The OpenGL graphics system: A specification (version 1.1)}, + author = {Segal, Mark and Akeley, Kurt}, + year = 1999 +} +@article{segura2018ethical, + title = {Ethical implications of user perceptions of wearable devices}, + author = {Segura Anaya, LH and Alsadoon, Abeer and Costadopoulos, Nectar and Prasad, PWC}, + year = 2018, + journal = {Science and engineering ethics}, + publisher = {Springer}, + volume = 24, + pages = {1--28} +} @inproceedings{seide2016cntk, - author = {Seide, Frank and Agarwal, Amit}, - booktitle = {Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining}, - pages = {2135--2135}, - title = {CNTK: Microsoft's open-source deep-learning toolkit}, - year = 2016} - + title = {CNTK: Microsoft's open-source deep-learning toolkit}, + author = {Seide, Frank and Agarwal, Amit}, + year = 2016, + booktitle = {Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining}, + pages = {2135--2135} +} +@inproceedings{selvaraju2017grad, + title = {Grad-cam: Visual explanations from deep networks via gradient-based localization}, + author = {Selvaraju, Ramprasaath R and Cogswell, Michael and Das, Abhishek and Vedantam, Ramakrishna and Parikh, Devi and Batra, Dhruv}, + year = 2017, + booktitle = {Proceedings of the IEEE international conference on computer vision}, + pages = {618--626} +} @misc{sevilla_compute_2022, - author = {Sevilla, Jaime and Heim, Lennart and Ho, Anson and Besiroglu, Tamay and Hobbhahn, Marius and Villalobos, Pablo}, - file = {Sevilla et al. - 2022 - Compute Trends Across Three Eras of Machine Learni.pdf:/Users/alex/Zotero/storage/24N9RZ72/Sevilla et al. - 2022 - Compute Trends Across Three Eras of Machine Learni.pdf:application/pdf}, - keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence, Computer Science - Computers and Society}, - language = {en}, - month = mar, - note = {arXiv:2202.05924 [cs]}, - publisher = {arXiv}, - title = {Compute {Trends} {Across} {Three} {Eras} of {Machine} {Learning}}, - url = {http://arxiv.org/abs/2202.05924}, - urldate = {2023-10-25}, - year = 2022, - Bdsk-Url-1 = {http://arxiv.org/abs/2202.05924}} - + title = {Compute {Trends} {Across} {Three} {Eras} of {Machine} {Learning}}, + author = {Sevilla, Jaime and Heim, Lennart and Ho, Anson and Besiroglu, Tamay and Hobbhahn, Marius and Villalobos, Pablo}, + year = 2022, + month = mar, + publisher = {arXiv}, + url = {http://arxiv.org/abs/2202.05924}, + urldate = {2023-10-25}, + note = {arXiv:2202.05924 [cs]}, + file = {Sevilla et al. - 2022 - Compute Trends Across Three Eras of Machine Learni.pdf:/Users/alex/Zotero/storage/24N9RZ72/Sevilla et al. - 2022 - Compute Trends Across Three Eras of Machine Learni.pdf:application/pdf}, + keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence, Computer Science - Computers and Society}, + language = {en}, + bdsk-url-1 = {http://arxiv.org/abs/2202.05924} +} @article{seyedzadeh2018machine, - author = {Seyedzadeh, Saleh and Rahimian, Farzad Pour and Glesk, Ivan and Roper, Marc}, - journal = {Visualization in Engineering}, - pages = {1--20}, - publisher = {Springer}, - title = {Machine learning for estimation of building energy consumption and performance: a review}, - volume = 6, - year = 2018} - + title = {Machine learning for estimation of building energy consumption and performance: a review}, + author = {Seyedzadeh, Saleh and Rahimian, Farzad Pour and Glesk, Ivan and Roper, Marc}, + year = 2018, + journal = {Visualization in Engineering}, + publisher = {Springer}, + volume = 6, + pages = {1--20} +} +@article{sgd, + title = {{A Stochastic Approximation Method}}, + author = {Herbert Robbins and Sutton Monro}, + year = 1951, + journal = {The Annals of Mathematical Statistics}, + url = {https://doi.org/10.1214/aoms/1177729586} +} +@article{shalev2017formal, + title = {On a formal model of safe and scalable self-driving cars}, + author = {Shalev-Shwartz, Shai and Shammah, Shaked and Shashua, Amnon}, + year = 2017, + journal = {arXiv preprint arXiv:1708.06374} +} @article{shamir1979share, - author = {Shamir, Adi}, - journal = {Communications of the ACM}, - number = 11, - pages = {612--613}, - publisher = {ACm New York, NY, USA}, - title = {How to share a secret}, - volume = 22, - year = 1979} - + title = {How to share a secret}, + author = {Shamir, Adi}, + year = 1979, + journal = {Communications of the ACM}, + publisher = {ACm New York, NY, USA}, + volume = 22, + number = 11, + pages = {612--613} +} +@article{shan2023prompt, + title = {Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models}, + author = {Shan, Shawn and Ding, Wenxin and Passananti, Josephine and Zheng, Haitao and Zhao, Ben Y}, + year = 2023, + journal = {arXiv preprint arXiv:2310.13828} +} +@article{shastri2021photonics, + title = {Photonics for artificial intelligence and neuromorphic computing}, + author = {Shastri, Bhavin J and Tait, Alexander N and Ferreira de Lima, Thomas and Pernice, Wolfram HP and Bhaskaran, Harish and Wright, C David and Prucnal, Paul R}, + year = 2021, + journal = {Nature Photonics}, + publisher = {Nature Publishing Group UK London}, + volume = 15, + number = 2, + pages = {102--114} +} @article{Sheng_Zhang_2019, - author = {Sheng, Victor S. and Zhang, Jing}, - doi = {10.1609/aaai.v33i01.33019837}, - journal = {Proceedings of the AAAI Conference on Artificial Intelligence}, - number = {01}, - pages = {9837--9843}, - title = {Machine learning with crowdsourcing: A brief summary of the past research and Future Directions}, - volume = 33, - year = 2019, - Bdsk-Url-1 = {https://doi.org/10.1609/aaai.v33i01.33019837}} - + title = {Machine learning with crowdsourcing: A brief summary of the past research and Future Directions}, + author = {Sheng, Victor S. and Zhang, Jing}, + year = 2019, + journal = {Proceedings of the AAAI Conference on Artificial Intelligence}, + volume = 33, + number = {01}, + pages = {9837--9843}, + doi = {10.1609/aaai.v33i01.33019837}, + bdsk-url-1 = {https://doi.org/10.1609/aaai.v33i01.33019837} +} @misc{Sheth_2022, - author = {Sheth, Dhruv}, - journal = {Hackster.io}, - month = {Mar}, - title = {Eletect - TinyML and IOT based Smart Wildlife Tracker}, - url = {https://www.hackster.io/dhruvsheth_/eletect-tinyml-and-iot-based-smart-wildlife-tracker-c03e5a}, - year = 2022, - Bdsk-Url-1 = {https://www.hackster.io/dhruvsheth_/eletect-tinyml-and-iot-based-smart-wildlife-tracker-c03e5a}} - + title = {Eletect - TinyML and IOT based Smart Wildlife Tracker}, + author = {Sheth, Dhruv}, + year = 2022, + month = {Mar}, + journal = {Hackster.io}, + url = {https://www.hackster.io/dhruvsheth_/eletect-tinyml-and-iot-based-smart-wildlife-tracker-c03e5a}, + bdsk-url-1 = {https://www.hackster.io/dhruvsheth_/eletect-tinyml-and-iot-based-smart-wildlife-tracker-c03e5a} +} @inproceedings{shi2022data, - author = {Shi, Hongrui and Radu, Valentin}, - booktitle = {Proceedings of the 2nd European Workshop on Machine Learning and Systems}, - pages = {72--78}, - title = {Data selection for efficient model update in federated learning}, - year = 2022} - + title = {Data selection for efficient model update in federated learning}, + author = {Shi, Hongrui and Radu, Valentin}, + year = 2022, + booktitle = {Proceedings of the 2nd European Workshop on Machine Learning and Systems}, + pages = {72--78} +} +@book{shneiderman2022human, + title = {Human-centered AI}, + author = {Shneiderman, Ben}, + year = 2022, + publisher = {Oxford University Press} +} +@inproceedings{shokri2017membership, + title = {Membership inference attacks against machine learning models}, + author = {Shokri, Reza and Stronati, Marco and Song, Congzheng and Shmatikov, Vitaly}, + year = 2017, + booktitle = {2017 IEEE symposium on security and privacy (SP)}, + pages = {3--18}, + organization = {IEEE} +} +@article{silvestro2022improving, + title = {Improving biodiversity protection through artificial intelligence}, + author = {Silvestro, Daniele and Goria, Stefano and Sterner, Thomas and Antonelli, Alexandre}, + year = 2022, + journal = {Nature sustainability}, + publisher = {Nature Publishing Group UK London}, + volume = 5, + number = 5, + pages = {415--424} +} +@inproceedings{skorobogatov2003optical, + title = {Optical fault induction attacks}, + author = {Skorobogatov, Sergei P and Anderson, Ross J}, + year = 2003, + booktitle = {Cryptographic Hardware and Embedded Systems-CHES 2002: 4th International Workshop Redwood Shores, CA, USA, August 13--15, 2002 Revised Papers 4}, + pages = {2--12}, + organization = {Springer} +} +@inproceedings{skorobogatov2009local, + title = {Local heating attacks on flash memory devices}, + author = {Skorobogatov, Sergei}, + year = 2009, + booktitle = {2009 IEEE International Workshop on Hardware-Oriented Security and Trust}, + pages = {1--6}, + organization = {IEEE} +} @article{smestad2023systematic, - author = {Smestad, Carl and Li, Jingyue}, - journal = {arXiv preprint arXiv:2306.04862}, - title = {A Systematic Literature Review on Client Selection in Federated Learning}, - year = 2023} - + title = {A Systematic Literature Review on Client Selection in Federated Learning}, + author = {Smestad, Carl and Li, Jingyue}, + year = 2023, + journal = {arXiv preprint arXiv:2306.04862} +} +@article{smilkov2017smoothgrad, + title = {Smoothgrad: removing noise by adding noise}, + author = {Smilkov, Daniel and Thorat, Nikhil and Kim, Been and Vi{\'e}gas, Fernanda and Wattenberg, Martin}, + year = 2017, + journal = {arXiv preprint arXiv:1706.03825} +} @misc{smoothquant, - abstract = {Large language models (LLMs) show excellent performance but are compute- and memory-intensive. Quantization can reduce memory and accelerate inference. However, existing methods cannot maintain accuracy and hardware efficiency at the same time. We propose SmoothQuant, a training-free, accuracy-preserving, and general-purpose post-training quantization (PTQ) solution to enable 8-bit weight, 8-bit activation (W8A8) quantization for LLMs. Based on the fact that weights are easy to quantize while activations are not, SmoothQuant smooths the activation outliers by offline migrating the quantization difficulty from activations to weights with a mathematically equivalent transformation. SmoothQuant enables an INT8 quantization of both weights and activations for all the matrix multiplications in LLMs, including OPT, BLOOM, GLM, MT-NLG, and LLaMA family. We demonstrate up to 1.56x speedup and 2x memory reduction for LLMs with negligible loss in accuracy. SmoothQuant enables serving 530B LLM within a single node. Our work offers a turn-key solution that reduces hardware costs and democratizes LLMs.}, - author = {Xiao and Lin, Seznec and Wu, Demouth and Han}, - doi = {10.48550/arXiv.2211.10438}, - title = {SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models}, - url = {https://arxiv.org/abs/2211.10438}, - urldate = {2023-06-05}, - year = 2023, - Bdsk-Url-1 = {https://arxiv.org/abs/2211.10438}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.2211.10438}} - + title = {SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models}, + author = {Xiao and Lin, Seznec and Wu, Demouth and Han}, + year = 2023, + doi = {10.48550/arXiv.2211.10438}, + url = {https://arxiv.org/abs/2211.10438}, + urldate = {2023-06-05}, + abstract = {Large language models (LLMs) show excellent performance but are compute- and memory-intensive. Quantization can reduce memory and accelerate inference. However, existing methods cannot maintain accuracy and hardware efficiency at the same time. We propose SmoothQuant, a training-free, accuracy-preserving, and general-purpose post-training quantization (PTQ) solution to enable 8-bit weight, 8-bit activation (W8A8) quantization for LLMs. Based on the fact that weights are easy to quantize while activations are not, SmoothQuant smooths the activation outliers by offline migrating the quantization difficulty from activations to weights with a mathematically equivalent transformation. SmoothQuant enables an INT8 quantization of both weights and activations for all the matrix multiplications in LLMs, including OPT, BLOOM, GLM, MT-NLG, and LLaMA family. We demonstrate up to 1.56x speedup and 2x memory reduction for LLMs with negligible loss in accuracy. SmoothQuant enables serving 530B LLM within a single node. Our work offers a turn-key solution that reduces hardware costs and democratizes LLMs.}, + bdsk-url-1 = {https://arxiv.org/abs/2211.10438}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.2211.10438} +} +@article{soufleri2023synthetic, + title = {Synthetic Dataset Generation for Privacy-Preserving Machine Learning}, + author = {Efstathia Soufleri and Gobinda Saha and Kaushik Roy}, + year = 2023, + journal = {arXiv preprint arXiv:2210.03205}, + date-added = {2023-11-22 19:26:18 -0500}, + date-modified = {2023-11-22 19:26:57 -0500} +} +@misc{strubell2019energy, + title = {Energy and Policy Considerations for Deep Learning in NLP}, + author = {Emma Strubell and Ananya Ganesh and Andrew McCallum}, + year = 2019, + eprint = {1906.02243}, + archiveprefix = {arXiv}, + primaryclass = {cs.CL} +} +@article{strubellEnergyPolicyConsiderations2019, + shorttitle = {Energy and Policy Considerations for Deep Learning in NLP}, + author = {Strubell, Emma and Ganesh, Ananya and McCallum, Andrew}, + year = 2019, + doi = {10.48550/ARXIV.1906.02243}, + url = {https://arxiv.org/abs/1906.02243}, + urldate = {2023-12-06}, + copyright = {arXiv.org perpetual, non-exclusive license}, + note = {Publisher: arXiv Version Number: 1}, + annote = {Other In the 57th Annual Meeting of the Association for Computational Linguistics (ACL). Florence, Italy. July 2019} +} +@inproceedings{suda2016throughput, + title = {Throughput-optimized OpenCL-based FPGA accelerator for large-scale convolutional neural networks}, + author = {Suda, Naveen and Chandra, Vikas and Dasika, Ganesh and Mohanty, Abinash and Ma, Yufei and Vrudhula, Sarma and Seo, Jae-sun and Cao, Yu}, + year = 2016, + booktitle = {Proceedings of the 2016 ACM/SIGDA international symposium on field-programmable gate arrays}, + pages = {16--25} +} @misc{surveyofquant, - abstract = {As soon as abstract mathematical computations were adapted to computation on digital computers, the problem of efficient representation, manipulation, and communication of the numerical values in those computations arose. Strongly related to the problem of numerical representation is the problem of quantization: in what manner should a set of continuous real-valued numbers be distributed over a fixed discrete set of numbers to minimize the number of bits required and also to maximize the accuracy of the attendant computations? This perennial problem of quantization is particularly relevant whenever memory and/or computational resources are severely restricted, and it has come to the forefront in recent years due to the remarkable performance of Neural Network models in computer vision, natural language processing, and related areas. Moving from floating-point representations to low-precision fixed integer values represented in four bits or less holds the potential to reduce the memory footprint and latency by a factor of 16x; and, in fact, reductions of 4x to 8x are often realized in practice in these applications. Thus, it is not surprising that quantization has emerged recently as an important and very active sub-area of research in the efficient implementation of computations associated with Neural Networks. In this article, we survey approaches to the problem of quantizing the numerical values in deep Neural Network computations, covering the advantages/disadvantages of current methods. With this survey and its organization, we hope to have presented a useful snapshot of the current research in quantization for Neural Networks and to have given an intelligent organization to ease the evaluation of future research in this area.}, - author = {Gholami and Kim, Dong and Yao, Mahoney and Keutzer}, - doi = {10.48550/arXiv.2103.13630}, - title = {A Survey of Quantization Methods for Efficient Neural Network Inference)}, - url = {https://arxiv.org/abs/2103.13630}, - urldate = {2021-06-21}, - year = 2021, - Bdsk-Url-1 = {https://arxiv.org/abs/2103.13630}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.2103.13630}} - + title = {A Survey of Quantization Methods for Efficient Neural Network Inference)}, + author = {Gholami and Kim, Dong and Yao, Mahoney and Keutzer}, + year = 2021, + doi = {10.48550/arXiv.2103.13630}, + url = {https://arxiv.org/abs/2103.13630}, + urldate = {2021-06-21}, + abstract = {As soon as abstract mathematical computations were adapted to computation on digital computers, the problem of efficient representation, manipulation, and communication of the numerical values in those computations arose. Strongly related to the problem of numerical representation is the problem of quantization: in what manner should a set of continuous real-valued numbers be distributed over a fixed discrete set of numbers to minimize the number of bits required and also to maximize the accuracy of the attendant computations? This perennial problem of quantization is particularly relevant whenever memory and/or computational resources are severely restricted, and it has come to the forefront in recent years due to the remarkable performance of Neural Network models in computer vision, natural language processing, and related areas. Moving from floating-point representations to low-precision fixed integer values represented in four bits or less holds the potential to reduce the memory footprint and latency by a factor of 16x; and, in fact, reductions of 4x to 8x are often realized in practice in these applications. Thus, it is not surprising that quantization has emerged recently as an important and very active sub-area of research in the efficient implementation of computations associated with Neural Networks. In this article, we survey approaches to the problem of quantizing the numerical values in deep Neural Network computations, covering the advantages/disadvantages of current methods. With this survey and its organization, we hope to have presented a useful snapshot of the current research in quantization for Neural Networks and to have given an intelligent organization to ease the evaluation of future research in this area.}, + bdsk-url-1 = {https://arxiv.org/abs/2103.13630}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.2103.13630} +} +@article{Sze2017-ak, + title = {Efficient processing of deep neural networks: A tutorial and survey}, + author = {Sze, Vivienne and Chen, Yu-Hsin and Yang, Tien-Ju and Emer, Joel}, + year = 2017, + month = mar, + copyright = {http://arxiv.org/licenses/nonexclusive-distrib/1.0/}, + abstract = {Deep neural networks (DNNs) are currently widely used for many artificial intelligence (AI) applications including computer vision, speech recognition, and robotics. While DNNs deliver state-of-the-art accuracy on many AI tasks, it comes at the cost of high computational complexity. Accordingly, techniques that enable efficient processing of DNNs to improve energy efficiency and throughput without sacrificing application accuracy or increasing hardware cost are critical to the wide deployment of DNNs in AI systems. This article aims to provide a comprehensive tutorial and survey about the recent advances towards the goal of enabling efficient processing of DNNs. Specifically, it will provide an overview of DNNs, discuss various hardware platforms and architectures that support DNNs, and highlight key trends in reducing the computation cost of DNNs either solely via hardware design changes or via joint hardware design and DNN algorithm changes. It will also summarize various development resources that enable researchers and practitioners to quickly get started in this field, and highlight important benchmarking metrics and design considerations that should be used for evaluating the rapidly growing number of DNN hardware designs, optionally including algorithmic co-designs, being proposed in academia and industry. The reader will take away the following concepts from this article: understand the key design considerations for DNNs; be able to evaluate different DNN hardware implementations with benchmarks and comparison metrics; understand the trade-offs between various hardware architectures and platforms; be able to evaluate the utility of various DNN design techniques for efficient processing; and understand recent implementation trends and opportunities.}, + archiveprefix = {arXiv}, + eprint = {1703.09039}, + primaryclass = {cs.CV} +} +@article{sze2017efficient, + title = {Efficient processing of deep neural networks: A tutorial and survey}, + author = {Sze, Vivienne and Chen, Yu-Hsin and Yang, Tien-Ju and Emer, Joel S}, + year = 2017, + journal = {Proceedings of the IEEE}, + publisher = {Ieee}, + volume = 105, + number = 12, + pages = {2295--2329} +} +@article{szegedy2013intriguing, + title = {Intriguing properties of neural networks}, + author = {Szegedy, Christian and Zaremba, Wojciech and Sutskever, Ilya and Bruna, Joan and Erhan, Dumitru and Goodfellow, Ian and Fergus, Rob}, + year = 2013, + journal = {arXiv preprint arXiv:1312.6199} +} @misc{tan_efficientnet_2020, - abstract = {Convolutional Neural Networks (ConvNets) are commonly developed at a fixed resource budget, and then scaled up for better accuracy if more resources are available. In this paper, we systematically study model scaling and identify that carefully balancing network depth, width, and resolution can lead to better performance. Based on this observation, we propose a new scaling method that uniformly scales all dimensions of depth/width/resolution using a simple yet highly effective compound coefficient. We demonstrate the effectiveness of this method on scaling up MobileNets and ResNet. To go even further, we use neural architecture search to design a new baseline network and scale it up to obtain a family of models, called EfficientNets, which achieve much better accuracy and efficiency than previous ConvNets. In particular, our EfficientNet-B7 achieves state-of-the-art 84.3\% top-1 accuracy on ImageNet, while being 8.4x smaller and 6.1x faster on inference than the best existing ConvNet. Our EfficientNets also transfer well and achieve state-of-the-art accuracy on CIFAR-100 (91.7\%), Flowers (98.8\%), and 3 other transfer learning datasets, with an order of magnitude fewer parameters. Source code is at https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet.}, - author = {Tan, Mingxing and Le, Quoc V.}, - doi = {10.48550/arXiv.1905.11946}, - file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/KISBF35I/Tan and Le - 2020 - EfficientNet Rethinking Model Scaling for Convolu.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/TUD4PH4M/1905.html:text/html}, - keywords = {Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning, Statistics - Machine Learning}, - month = sep, - note = {arXiv:1905.11946 [cs, stat]}, - publisher = {arXiv}, - shorttitle = {{EfficientNet}}, - title = {{EfficientNet}: {Rethinking} {Model} {Scaling} for {Convolutional} {Neural} {Networks}}, - url = {http://arxiv.org/abs/1905.11946}, - urldate = {2023-10-20}, - year = 2020, - Bdsk-Url-1 = {http://arxiv.org/abs/1905.11946}, - Bdsk-Url-2 = {https://doi.org/10.48550/arXiv.1905.11946}} - + title = {{EfficientNet}: {Rethinking} {Model} {Scaling} for {Convolutional} {Neural} {Networks}}, + shorttitle = {{EfficientNet}}, + author = {Tan, Mingxing and Le, Quoc V.}, + year = 2020, + month = sep, + publisher = {arXiv}, + doi = {10.48550/arXiv.1905.11946}, + url = {http://arxiv.org/abs/1905.11946}, + urldate = {2023-10-20}, + note = {arXiv:1905.11946 [cs, stat]}, + abstract = {Convolutional Neural Networks (ConvNets) are commonly developed at a fixed resource budget, and then scaled up for better accuracy if more resources are available. In this paper, we systematically study model scaling and identify that carefully balancing network depth, width, and resolution can lead to better performance. Based on this observation, we propose a new scaling method that uniformly scales all dimensions of depth/width/resolution using a simple yet highly effective compound coefficient. We demonstrate the effectiveness of this method on scaling up MobileNets and ResNet. To go even further, we use neural architecture search to design a new baseline network and scale it up to obtain a family of models, called EfficientNets, which achieve much better accuracy and efficiency than previous ConvNets. In particular, our EfficientNet-B7 achieves state-of-the-art 84.3\% top-1 accuracy on ImageNet, while being 8.4x smaller and 6.1x faster on inference than the best existing ConvNet. Our EfficientNets also transfer well and achieve state-of-the-art accuracy on CIFAR-100 (91.7\%), Flowers (98.8\%), and 3 other transfer learning datasets, with an order of magnitude fewer parameters. Source code is at https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet.}, + file = {arXiv Fulltext PDF:/Users/jeffreyma/Zotero/storage/KISBF35I/Tan and Le - 2020 - EfficientNet Rethinking Model Scaling for Convolu.pdf:application/pdf;arXiv.org Snapshot:/Users/jeffreyma/Zotero/storage/TUD4PH4M/1905.html:text/html}, + keywords = {Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning, Statistics - Machine Learning}, + bdsk-url-1 = {http://arxiv.org/abs/1905.11946}, + bdsk-url-2 = {https://doi.org/10.48550/arXiv.1905.11946} +} @inproceedings{tan2019mnasnet, - author = {Tan, Mingxing and Chen, Bo and Pang, Ruoming and Vasudevan, Vijay and Sandler, Mark and Howard, Andrew and Le, Quoc V}, - booktitle = {Proceedings of the IEEE/CVF conference on computer vision and pattern recognition}, - pages = {2820--2828}, - title = {Mnasnet: Platform-aware neural architecture search for mobile}, - year = 2019} - + title = {Mnasnet: Platform-aware neural architecture search for mobile}, + author = {Tan, Mingxing and Chen, Bo and Pang, Ruoming and Vasudevan, Vijay and Sandler, Mark and Howard, Andrew and Le, Quoc V}, + year = 2019, + booktitle = {Proceedings of the IEEE/CVF conference on computer vision and pattern recognition}, + pages = {2820--2828} +} @misc{tan2020efficientnet, + title = {EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks}, + author = {Mingxing Tan and Quoc V. Le}, + year = 2020, archiveprefix = {arXiv}, - author = {Mingxing Tan and Quoc V. Le}, - eprint = {1905.11946}, - primaryclass = {cs.LG}, - title = {EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks}, - year = 2020} - + eprint = {1905.11946}, + primaryclass = {cs.LG} +} +@article{tang2022soft, + title = {Soft bioelectronics for cardiac interfaces}, + author = {Tang, Xin and He, Yichun and Liu, Jia}, + year = 2022, + journal = {Biophysics Reviews}, + publisher = {AIP Publishing}, + volume = 3, + number = 1 +} +@article{tang2023flexible, + title = {Flexible brain--computer interfaces}, + author = {Tang, Xin and Shen, Hao and Zhao, Siyuan and Li, Na and Liu, Jia}, + year = 2023, + journal = {Nature Electronics}, + publisher = {Nature Publishing Group UK London}, + volume = 6, + number = 2, + pages = {109--118} +} +@article{tarun2023deep, + title = {Deep Regression Unlearning}, + author = {Ayush K Tarun and Vikram S Chundawat and Murari Mandal and Mohan Kankanhalli}, + year = 2023, + journal = {arXiv preprint arXiv:2210.08196}, + date-added = {2023-11-22 19:20:59 -0500}, + date-modified = {2023-11-22 19:21:59 -0500} +} @misc{Team_2023, - author = {Team, Snorkel}, - journal = {Snorkel AI}, - month = {Aug}, - title = {Data-centric AI for the Enterprise}, - url = {https://snorkel.ai/}, - year = 2023, - Bdsk-Url-1 = {https://snorkel.ai/}} - + title = {Data-centric AI for the Enterprise}, + author = {Team, Snorkel}, + year = 2023, + month = {Aug}, + journal = {Snorkel AI}, + url = {https://snorkel.ai/}, + bdsk-url-1 = {https://snorkel.ai/} +} @misc{Thefutur92:online, - author = {ARM.com}, - howpublished = {\url{https://www.arm.com/company/news/2023/02/arm-announces-q3-fy22-results}}, - note = {(Accessed on 09/16/2023)}, - title = {The future is being built on Arm: Market diversification continues to drive strong royalty and licensing growth as ecosystem reaches quarter of a trillion chips milestone -- Arm{\textregistered}}} - + title = {The future is being built on Arm: Market diversification continues to drive strong royalty and licensing growth as ecosystem reaches quarter of a trillion chips milestone -- Arm{\textregistered}}, + author = {ARM.com}, + note = {(Accessed on 09/16/2023)}, + howpublished = {\url{https://www.arm.com/company/news/2023/02/arm-announces-q3-fy22-results}} +} @misc{threefloat, - author = {Google}, - title = {Three Floating Point Formats}, - url = {https://storage.googleapis.com/gweb-cloudblog-publish/images/Three_floating-point_formats.max-624x261.png}, - urldate = {2023-10-20}, - year = 2023, - Bdsk-Url-1 = {https://storage.googleapis.com/gweb-cloudblog-publish/images/Three_floating-point_formats.max-624x261.png}} - + title = {Three Floating Point Formats}, + author = {Google}, + year = 2023, + url = {https://storage.googleapis.com/gweb-cloudblog-publish/images/Three_floating-point_formats.max-624x261.png}, + urldate = {2023-10-20}, + bdsk-url-1 = {https://storage.googleapis.com/gweb-cloudblog-publish/images/Three_floating-point_formats.max-624x261.png} +} +@article{tillFishDieoffsAre2019, + title = {Fish die-offs are concurrent with thermal extremes in north temperate lakes}, + author = {Till, Aaron and Rypel, Andrew L and Bray, Andrew and Fey, Samuel B}, + year = 2019, + journal = {Nature Climate Change}, + publisher = {Nature Publishing Group UK London}, + volume = 9, + number = 8, + pages = {637--641} +} @article{tirtalistyani2022indonesia, - author = {Tirtalistyani, Rose and Murtiningrum, Murtiningrum and Kanwar, Rameshwar S}, - journal = {Sustainability}, - number = 19, - pages = 12477, - publisher = {MDPI}, - title = {Indonesia rice irrigation system: Time for innovation}, - volume = 14, - year = 2022} - + title = {Indonesia rice irrigation system: Time for innovation}, + author = {Tirtalistyani, Rose and Murtiningrum, Murtiningrum and Kanwar, Rameshwar S}, + year = 2022, + journal = {Sustainability}, + publisher = {MDPI}, + volume = 14, + number = 19, + pages = 12477 +} @inproceedings{tokui2015chainer, - author = {Tokui, Seiya and Oono, Kenta and Hido, Shohei and Clayton, Justin}, - booktitle = {Proceedings of workshop on machine learning systems (LearningSys) in the twenty-ninth annual conference on neural information processing systems (NIPS)}, - pages = {1--6}, - title = {Chainer: a next-generation open source framework for deep learning}, - volume = 5, - year = 2015} - + title = {Chainer: a next-generation open source framework for deep learning}, + author = {Tokui, Seiya and Oono, Kenta and Hido, Shohei and Clayton, Justin}, + year = 2015, + booktitle = {Proceedings of workshop on machine learning systems (LearningSys) in the twenty-ninth annual conference on neural information processing systems (NIPS)}, + volume = 5, + pages = {1--6} +} +@inproceedings{tramer2019adversarial, + title = {Adversarial: Perceptual ad blocking meets adversarial machine learning}, + author = {Tram{\`e}r, Florian and Dupr{\'e}, Pascal and Rusak, Gili and Pellegrino, Giancarlo and Boneh, Dan}, + year = 2019, + booktitle = {Proceedings of the 2019 ACM SIGSAC conference on computer and communications security}, + pages = {2005--2021} +} +@article{uddin_energy_2012, + title = {Energy efficiency and low carbon enabler green {IT} framework for data centers considering green metrics}, + author = {Uddin, Mueen and Rahman, Azizah Abdul}, + year = 2012, + journal = {Renewable and Sustainable Energy Reviews}, + volume = 16, + number = 6, + pages = {4078--4094}, + doi = {https://doi.org/10.1016/j.rser.2012.03.014}, + issn = {1364-0321}, + url = {https://www.sciencedirect.com/science/article/pii/S1364032112001979}, + abstract = {The increasing demand for storage, networking and computation has driven intensification of large complex data centers that run many of today's Internet, financial, commercial and business applications. A data center comprises of many thousands of servers and can use as much energy as small city. Massive amount of computation power is required to drive and run these server farms resulting in many challenging like huge energy consumptions, emission of green house gases, backups and recovery; This paper proposes energy efficiency and low carbon enabler green IT framework for these large and complex server farms to save consumption of electricity and reduce the emission of green house gases to lower the effects of global warming. The framework uses latest energy saving techniques like virtualization, cloud computing and green metrics to achieve greener data centers. It comprises of five phase to properly implement green IT techniques to achieve green data centers. The proposed framework seamlessly divides data center components into different resource pools and then applies green metrics like Power Usage Effectiveness, Data Center Effectiveness and Carbon Emission Calculator to measure performance of individual components so that benchmarking values can be achieved and set as standard to be followed by data centers.}, + keywords = {Energy efficiency, Energy efficient data centers, Global warming, Green IT, Green IT framework, Green metrics, Virtualization} +} +@book{un_world_economic_forum_2019, + title = {A New Circular Vision for Electronics, Time for a Global Reboot}, + author = {UN and World Economic Forum}, + year = 2019, + month = {Dec}, + publisher = {PACE - Platform for Accelerating the Circular Economy}, + url = {https://www3.weforum.org/docs/WEF_A_New_Circular_Vision_for_Electronics.pdf} +} +@techreport{uptime, + title = {Uptime Institute Global Data Center Survey 2022}, + author = {Jacqueline Davis, Research Analyst Daniel Bizo, Research Director Andy Lawrence, Executive Director of Research Owen Rogers, Research Director for Cloud Computing Max Smolaks, Research Analyst}, + year = 2022, + institution = {Uptime Institute} +} +@techreport{USA_energy, + title = {United States Data Center Energy Usage Report}, + author = {Shehabi, Arman, Sarah Josephine Smith, Dale A. Sartor, Richard E. Brown, Magnus Herrlin, Jonathan G. Koomey, Eric R. Masanet, Nathaniel Horner, Inês Lima Azevedo, William Lintner}, + year = 2016, + institution = {Berkeley Laboratory} +} +@article{USA_footprint, + title = {The environmental footprint of data centers in the United States}, + author = {Siddik, Md Abu Bakar and Shehabi, Arman and Marston, Landon}, + year = 2021, + month = may, + journal = {Environmental Research Letters}, + publisher = {IOP Publishing}, + volume = 16, + number = 6, + pages = {064017}, + doi = {10.1088/1748-9326/abfba1}, + issn = {1748-9326}, + url = {http://dx.doi.org/10.1088/1748-9326/abfba1} +} @article{van_de_ven_three_2022, - author = {Van De Ven, Gido M. and Tuytelaars, Tinne and Tolias, Andreas S.}, - doi = {10.1038/s42256-022-00568-3}, - file = {Van De Ven et al. - 2022 - Three types of incremental learning.pdf:/Users/alex/Zotero/storage/5ZAHXMQN/Van De Ven et al. - 2022 - Three types of incremental learning.pdf:application/pdf}, - issn = {2522-5839}, - journal = {Nature Machine Intelligence}, - language = {en}, - month = dec, - number = 12, - pages = {1185--1197}, - title = {Three types of incremental learning}, - url = {https://www.nature.com/articles/s42256-022-00568-3}, - urldate = {2023-10-26}, - volume = 4, - year = 2022, - Bdsk-Url-1 = {https://www.nature.com/articles/s42256-022-00568-3}, - Bdsk-Url-2 = {https://doi.org/10.1038/s42256-022-00568-3}} - + title = {Three types of incremental learning}, + author = {Van De Ven, Gido M. and Tuytelaars, Tinne and Tolias, Andreas S.}, + year = 2022, + month = dec, + journal = {Nature Machine Intelligence}, + volume = 4, + number = 12, + pages = {1185--1197}, + doi = {10.1038/s42256-022-00568-3}, + issn = {2522-5839}, + url = {https://www.nature.com/articles/s42256-022-00568-3}, + urldate = {2023-10-26}, + file = {Van De Ven et al. - 2022 - Three types of incremental learning.pdf:/Users/alex/Zotero/storage/5ZAHXMQN/Van De Ven et al. - 2022 - Three types of incremental learning.pdf:application/pdf}, + language = {en}, + bdsk-url-1 = {https://www.nature.com/articles/s42256-022-00568-3}, + bdsk-url-2 = {https://doi.org/10.1038/s42256-022-00568-3} +} @article{vaswani2017attention, - author = {Vaswani, Ashish and Shazeer, Noam and Parmar, Niki and Uszkoreit, Jakob and Jones, Llion and Gomez, Aidan N and Kaiser, {\L}ukasz and Polosukhin, Illia}, - journal = {Advances in neural information processing systems}, - title = {Attention is all you need}, - volume = 30, - year = 2017} - + title = {Attention is all you need}, + author = {Vaswani, Ashish and Shazeer, Noam and Parmar, Niki and Uszkoreit, Jakob and Jones, Llion and Gomez, Aidan N and Kaiser, {\L}ukasz and Polosukhin, Illia}, + year = 2017, + journal = {Advances in neural information processing systems}, + volume = 30 +} @misc{Vectorbo78:online, - howpublished = {\url{https://www.who.int/news-room/fact-sheets/detail/vector-borne-diseases}}, - note = {(Accessed on 10/17/2023)}, - title = {Vector-borne diseases}} - + title = {Vector-borne diseases}, + note = {(Accessed on 10/17/2023)}, + howpublished = {\url{https://www.who.int/news-room/fact-sheets/detail/vector-borne-diseases}} +} @misc{Verma_2022, - author = {Verma, Team Dual_Boot: Swapnil}, - journal = {Hackster.io}, - month = {Mar}, - title = {Elephant AI}, - url = {https://www.hackster.io/dual_boot/elephant-ai-ba71e9}, - year = 2022, - Bdsk-Url-1 = {https://www.hackster.io/dual_boot/elephant-ai-ba71e9}} - + title = {Elephant AI}, + author = {Verma, Team Dual_Boot: Swapnil}, + year = 2022, + month = {Mar}, + journal = {Hackster.io}, + url = {https://www.hackster.io/dual_boot/elephant-ai-ba71e9}, + bdsk-url-1 = {https://www.hackster.io/dual_boot/elephant-ai-ba71e9} +} +@article{verma2019memory, + title = {In-memory computing: Advances and prospects}, + author = {Verma, Naveen and Jia, Hongyang and Valavi, Hossein and Tang, Yinqi and Ozatay, Murat and Chen, Lung-Yen and Zhang, Bonan and Deaville, Peter}, + year = 2019, + journal = {IEEE Solid-State Circuits Magazine}, + publisher = {IEEE}, + volume = 11, + number = 3, + pages = {43--55} +} @misc{villalobos_machine_2022, - author = {Villalobos, Pablo and Sevilla, Jaime and Besiroglu, Tamay and Heim, Lennart and Ho, Anson and Hobbhahn, Marius}, - file = {Villalobos et al. - 2022 - Machine Learning Model Sizes and the Parameter Gap.pdf:/Users/alex/Zotero/storage/WW69A82B/Villalobos et al. - 2022 - Machine Learning Model Sizes and the Parameter Gap.pdf:application/pdf}, - keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence, Computer Science - Computers and Society, Computer Science - Computation and Language}, - language = {en}, - month = jul, - note = {arXiv:2207.02852 [cs]}, - publisher = {arXiv}, - title = {Machine {Learning} {Model} {Sizes} and the {Parameter} {Gap}}, - url = {http://arxiv.org/abs/2207.02852}, - urldate = {2023-10-25}, - year = 2022, - Bdsk-Url-1 = {http://arxiv.org/abs/2207.02852}} - + title = {Machine {Learning} {Model} {Sizes} and the {Parameter} {Gap}}, + author = {Villalobos, Pablo and Sevilla, Jaime and Besiroglu, Tamay and Heim, Lennart and Ho, Anson and Hobbhahn, Marius}, + year = 2022, + month = jul, + publisher = {arXiv}, + url = {http://arxiv.org/abs/2207.02852}, + urldate = {2023-10-25}, + note = {arXiv:2207.02852 [cs]}, + file = {Villalobos et al. - 2022 - Machine Learning Model Sizes and the Parameter Gap.pdf:/Users/alex/Zotero/storage/WW69A82B/Villalobos et al. - 2022 - Machine Learning Model Sizes and the Parameter Gap.pdf:application/pdf}, + keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence, Computer Science - Computers and Society, Computer Science - Computation and Language}, + language = {en}, + bdsk-url-1 = {http://arxiv.org/abs/2207.02852} +} @misc{villalobos_trends_2022, - author = {Villalobos, Pablo and Ho, Anson}, - journal = {Epoch AI}, - month = sep, - title = {Trends in {Training} {Dataset} {Sizes}}, - url = {https://epochai.org/blog/trends-in-training-dataset-sizes}, - year = 2022, - Bdsk-Url-1 = {https://epochai.org/blog/trends-in-training-dataset-sizes}} - + title = {Trends in {Training} {Dataset} {Sizes}}, + author = {Villalobos, Pablo and Ho, Anson}, + year = 2022, + month = sep, + journal = {Epoch AI}, + url = {https://epochai.org/blog/trends-in-training-dataset-sizes}, + bdsk-url-1 = {https://epochai.org/blog/trends-in-training-dataset-sizes} +} @misc{VinBrain, - journal = {VinBrain}, - url = {https://vinbrain.net/aiscaler}, - Bdsk-Url-1 = {https://vinbrain.net/aiscaler}} - + journal = {VinBrain}, + url = {https://vinbrain.net/aiscaler}, + bdsk-url-1 = {https://vinbrain.net/aiscaler} +} @article{vinuesa2020role, - author = {Vinuesa, Ricardo and Azizpour, Hossein and Leite, Iolanda and Balaam, Madeline and Dignum, Virginia and Domisch, Sami and Fell{\"a}nder, Anna and Langhans, Simone Daniela and Tegmark, Max and Fuso Nerini, Francesco}, - journal = {Nature communications}, - number = 1, - pages = {1--10}, - publisher = {Nature Publishing Group}, - title = {The role of artificial intelligence in achieving the Sustainable Development Goals}, - volume = 11, - year = 2020} - + title = {The role of artificial intelligence in achieving the Sustainable Development Goals}, + author = {Vinuesa, Ricardo and Azizpour, Hossein and Leite, Iolanda and Balaam, Madeline and Dignum, Virginia and Domisch, Sami and Fell{\"a}nder, Anna and Langhans, Simone Daniela and Tegmark, Max and Fuso Nerini, Francesco}, + year = 2020, + journal = {Nature communications}, + publisher = {Nature Publishing Group}, + volume = 11, + number = 1, + pages = {1--10} +} +@article{Vivet2021, + title = {IntAct: A 96-Core Processor With Six Chiplets 3D-Stacked on an Active Interposer With Distributed Interconnects and Integrated Power Management}, + author = {Vivet, Pascal and Guthmuller, Eric and Thonnart, Yvain and Pillonnet, Gael and Fuguet, C{\'e}sar and Miro-Panades, Ivan and Moritz, Guillaume and Durupt, Jean and Bernard, Christian and Varreau, Didier and Pontes, Julian and Thuries, S{\'e}bastien and Coriat, David and Harrand, Michel and Dutoit, Denis and Lattard, Didier and Arnaud, Lucile and Charbonnier, Jean and Coudrain, Perceval and Garnier, Arnaud and Berger, Fr{\'e}d{\'e}ric and Gueugnot, Alain and Greiner, Alain and Meunier, Quentin L. and Farcy, Alexis and Arriordaz, Alexandre and Ch{\'e}ramy, S{\'e}verine and Clermidy, Fabien}, + year = 2021, + journal = {IEEE Journal of Solid-State Circuits}, + volume = 56, + number = 1, + pages = {79--97}, + doi = {10.1109/JSSC.2020.3036341}, + bdsk-url-1 = {https://doi.org/10.1109/JSSC.2020.3036341} +} +@article{wachter2017counterfactual, + title = {Counterfactual explanations without opening the black box: Automated decisions and the GDPR}, + author = {Wachter, Sandra and Mittelstadt, Brent and Russell, Chris}, + year = 2017, + journal = {Harv. JL \& Tech.}, + publisher = {HeinOnline}, + volume = 31, + pages = 841 +} +@article{wald1987semiconductor, + title = {Semiconductor manufacturing: an introduction to processes and hazards}, + author = {Wald, Peter H and Jones, Jeffrey R}, + year = 1987, + journal = {American journal of industrial medicine}, + publisher = {Wiley Online Library}, + volume = 11, + number = 2, + pages = {203--221} +} +@article{waldSemiconductorManufacturingIntroduction1987, + title = {Semiconductor manufacturing: an introduction to processes and hazards}, + author = {Wald, Peter H and Jones, Jeffrey R}, + year = 1987, + journal = {American journal of industrial medicine}, + publisher = {Wiley Online Library}, + volume = 11, + number = 2, + pages = {203--221} +} +@inproceedings{wang2020apq, + title = {APQ: Joint Search for Network Architecture, Pruning and Quantization Policy}, + author = {Wang, Tianzhe and Wang, Kuan and Cai, Han and Lin, Ji and Liu, Zhijian and Wang, Hanrui and Lin, Yujun and Han, Song}, + year = 2020, + booktitle = {2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, + volume = {}, + number = {}, + pages = {2075--2084}, + doi = {10.1109/CVPR42600.2020.00215}, + bdsk-url-1 = {https://doi.org/10.1109/CVPR42600.2020.00215} +} +@article{wang2022interpretability, + title = {Interpretability in the wild: a circuit for indirect object identification in gpt-2 small}, + author = {Wang, Kevin and Variengien, Alexandre and Conmy, Arthur and Shlegeris, Buck and Steinhardt, Jacob}, + year = 2022, + journal = {arXiv preprint arXiv:2211.00593} +} @article{warden2018speech, - author = {Warden, Pete}, - journal = {arXiv preprint arXiv:1804.03209}, - title = {Speech commands: A dataset for limited-vocabulary speech recognition}, - year = 2018} - + title = {Speech commands: A dataset for limited-vocabulary speech recognition}, + author = {Warden, Pete}, + year = 2018, + journal = {arXiv preprint arXiv:1804.03209} +} @book{warden2019tinyml, - author = {Warden, Pete and Situnayake, Daniel}, - publisher = {O'Reilly Media}, - title = {Tinyml: Machine learning with tensorflow lite on arduino and ultra-low-power microcontrollers}, - year = 2019} - + title = {Tinyml: Machine learning with tensorflow lite on arduino and ultra-low-power microcontrollers}, + author = {Warden, Pete and Situnayake, Daniel}, + year = 2019, + publisher = {O'Reilly Media} +} +@article{wearableinsulin, + title = {Wearable Insulin Biosensors for Diabetes Management: Advances and Challenges}, + author = {Psoma, Sotiria D. and Kanthou, Chryso}, + year = 2023, + journal = {Biosensors}, + volume = 13, + number = 7, + doi = {10.3390/bios13070719}, + issn = {2079-6374}, + url = {https://www.mdpi.com/2079-6374/13/7/719}, + article-number = 719, + pubmedid = 37504117, + bdsk-url-1 = {https://www.mdpi.com/2079-6374/13/7/719}, + bdsk-url-2 = {https://doi.org/10.3390/bios13070719} +} +@techreport{weforum, + title = {Net-Zero Challenge: The supply chain opportunity}, + author = {Anthony Hobley, Dominic Waughray, Jens Burchardt, Michel Frédeau, Miranda Hadfield, Patrick Herhold, Henri Humpert, Christine O’Brien, Cornelius Pieper, Daniel Weise}, + year = 2021, + month = {jan}, + institution = {World Economic Forum, Boston Consulting Group} +} +@book{weik_survey_1955, + title = {A {Survey} of {Domestic} {Electronic} {Digital} {Computing} {Systems}}, + author = {Weik, Martin H.}, + year = 1955, + publisher = {Ballistic Research Laboratories}, + language = {en} +} @article{weiss_survey_2016, - author = {Weiss, Karl and Khoshgoftaar, Taghi M. and Wang, DingDing}, - doi = {10.1186/s40537-016-0043-6}, - file = {Weiss et al. - 2016 - A survey of transfer learning.pdf:/Users/alex/Zotero/storage/3FN2Y6EA/Weiss et al. - 2016 - A survey of transfer learning.pdf:application/pdf}, - issn = {2196-1115}, - journal = {Journal of Big Data}, - language = {en}, - month = dec, - number = 1, - pages = 9, - title = {A survey of transfer learning}, - url = {http://journalofbigdata.springeropen.com/articles/10.1186/s40537-016-0043-6}, - urldate = {2023-10-25}, - volume = 3, - year = 2016, - Bdsk-Url-1 = {http://journalofbigdata.springeropen.com/articles/10.1186/s40537-016-0043-6}, - Bdsk-Url-2 = {https://doi.org/10.1186/s40537-016-0043-6}} - + title = {A survey of transfer learning}, + author = {Weiss, Karl and Khoshgoftaar, Taghi M. and Wang, DingDing}, + year = 2016, + month = dec, + journal = {Journal of Big Data}, + volume = 3, + number = 1, + pages = 9, + doi = {10.1186/s40537-016-0043-6}, + issn = {2196-1115}, + url = {http://journalofbigdata.springeropen.com/articles/10.1186/s40537-016-0043-6}, + urldate = {2023-10-25}, + file = {Weiss et al. - 2016 - A survey of transfer learning.pdf:/Users/alex/Zotero/storage/3FN2Y6EA/Weiss et al. - 2016 - A survey of transfer learning.pdf:application/pdf}, + language = {en}, + bdsk-url-1 = {http://journalofbigdata.springeropen.com/articles/10.1186/s40537-016-0043-6}, + bdsk-url-2 = {https://doi.org/10.1186/s40537-016-0043-6} +} +@article{wiener1960some, + title = {Some Moral and Technical Consequences of Automation: As machines learn they may develop unforeseen strategies at rates that baffle their programmers.}, + author = {Wiener, Norbert}, + year = 1960, + journal = {Science}, + publisher = {American Association for the Advancement of Science}, + volume = 131, + number = 3410, + pages = {1355--1358} +} +@article{wong2012metal, + title = {Metal--oxide RRAM}, + author = {Wong, H-S Philip and Lee, Heng-Yuan and Yu, Shimeng and Chen, Yu-Sheng and Wu, Yi and Chen, Pang-Shiu and Lee, Byoungil and Chen, Frederick T and Tsai, Ming-Jinn}, + year = 2012, + journal = {Proceedings of the IEEE}, + publisher = {IEEE}, + volume = 100, + number = 6, + pages = {1951--1970} +} +@inproceedings{wong2018provable, + title = {Provable defenses against adversarial examples via the convex outer adversarial polytope}, + author = {Wong, Eric and Kolter, Zico}, + year = 2018, + booktitle = {International conference on machine learning}, + pages = {5286--5295}, + organization = {PMLR} +} @inproceedings{wu2019fbnet, - author = {Wu, Bichen and Dai, Xiaoliang and Zhang, Peizhao and Wang, Yanghan and Sun, Fei and Wu, Yiming and Tian, Yuandong and Vajda, Peter and Jia, Yangqing and Keutzer, Kurt}, - booktitle = {Proceedings of the IEEE/CVF conference on computer vision and pattern recognition}, - pages = {10734--10742}, - title = {Fbnet: Hardware-aware efficient convnet design via differentiable neural architecture search}, - year = 2019} - + title = {Fbnet: Hardware-aware efficient convnet design via differentiable neural architecture search}, + author = {Wu, Bichen and Dai, Xiaoliang and Zhang, Peizhao and Wang, Yanghan and Sun, Fei and Wu, Yiming and Tian, Yuandong and Vajda, Peter and Jia, Yangqing and Keutzer, Kurt}, + year = 2019, + booktitle = {Proceedings of the IEEE/CVF conference on computer vision and pattern recognition}, + pages = {10734--10742} +} @article{wu2022sustainable, - author = {Wu, Carole-Jean and Raghavendra, Ramya and Gupta, Udit and Acun, Bilge and Ardalani, Newsha and Maeng, Kiwan and Chang, Gloria and Aga, Fiona and Huang, Jinshi and Bai, Charles and others}, - journal = {Proceedings of Machine Learning and Systems}, - pages = {795--813}, - title = {Sustainable ai: Environmental implications, challenges and opportunities}, - volume = 4, - year = 2022} - + title = {Sustainable ai: Environmental implications, challenges and opportunities}, + author = {Wu, Carole-Jean and Raghavendra, Ramya and Gupta, Udit and Acun, Bilge and Ardalani, Newsha and Maeng, Kiwan and Chang, Gloria and Aga, Fiona and Huang, Jinshi and Bai, Charles and others}, + year = 2022, + journal = {Proceedings of Machine Learning and Systems}, + volume = 4, + pages = {795--813} +} +@inproceedings{xavier, + title = {Understanding the difficulty of training deep feedforward neural networks}, + author = {Glorot, Xavier and Bengio, Yoshua}, + year = 2010, + booktitle = {Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics}, + url = {https://proceedings.mlr.press/v9/glorot10a.html} +} @inproceedings{xie2020adversarial, - author = {Xie, Cihang and Tan, Mingxing and Gong, Boqing and Wang, Jiang and Yuille, Alan L and Le, Quoc V}, - booktitle = {Proceedings of the IEEE/CVF conference on computer vision and pattern recognition}, - pages = {819--828}, - title = {Adversarial examples improve image recognition}, - year = 2020} - + title = {Adversarial examples improve image recognition}, + author = {Xie, Cihang and Tan, Mingxing and Gong, Boqing and Wang, Jiang and Yuille, Alan L and Le, Quoc V}, + year = 2020, + booktitle = {Proceedings of the IEEE/CVF conference on computer vision and pattern recognition}, + pages = {819--828} +} +@article{xiong_mri-based_2021, + title = {{MRI}-based brain tumor segmentation using {FPGA}-accelerated neural network}, + author = {Xiong, Siyu and Wu, Guoqing and Fan, Xitian and Feng, Xuan and Huang, Zhongcheng and Cao, Wei and Zhou, Xuegong and Ding, Shijin and Yu, Jinhua and Wang, Lingli and Shi, Zhifeng}, + year = 2021, + month = sep, + journal = {BMC Bioinformatics}, + volume = 22, + number = 1, + pages = 421, + doi = {10.1186/s12859-021-04347-6}, + issn = {1471-2105}, + url = {https://doi.org/10.1186/s12859-021-04347-6}, + urldate = {2023-11-07}, + abstract = {Brain tumor segmentation is a challenging problem in medical image processing and analysis. It is a very time-consuming and error-prone task. In order to reduce the burden on physicians and improve the segmentation accuracy, the computer-aided detection (CAD) systems need to be developed. Due to the powerful feature learning ability of the deep learning technology, many deep learning-based methods have been applied to the brain tumor segmentation CAD systems and achieved satisfactory accuracy. However, deep learning neural networks have high computational complexity, and the brain tumor segmentation process consumes significant time. Therefore, in order to achieve the high segmentation accuracy of brain tumors and obtain the segmentation results efficiently, it is very demanding to speed up the segmentation process of brain tumors.}, + keywords = {Brain tumor segmatation, FPGA acceleration, Neural network}, + bdsk-url-1 = {https://doi.org/10.1186/s12859-021-04347-6} +} +@article{xiu2019time, + title = {Time Moore: Exploiting Moore's Law from the perspective of time}, + author = {Xiu, Liming}, + year = 2019, + journal = {IEEE Solid-State Circuits Magazine}, + publisher = {IEEE}, + volume = 11, + number = 1, + pages = {39--55} +} @article{xu2018alternating, - author = {Xu, Chen and Yao, Jianqiang and Lin, Zhouchen and Ou, Wenwu and Cao, Yuanbin and Wang, Zhirong and Zha, Hongbin}, - journal = {arXiv preprint arXiv:1802.00150}, - title = {Alternating multi-bit quantization for recurrent neural networks}, - year = 2018} - + title = {Alternating multi-bit quantization for recurrent neural networks}, + author = {Xu, Chen and Yao, Jianqiang and Lin, Zhouchen and Ou, Wenwu and Cao, Yuanbin and Wang, Zhirong and Zha, Hongbin}, + year = 2018, + journal = {arXiv preprint arXiv:1802.00150} +} @article{xu2023demystifying, - author = {Xu, Hu and Xie, Saining and Tan, Xiaoqing Ellen and Huang, Po-Yao and Howes, Russell and Sharma, Vasu and Li, Shang-Wen and Ghosh, Gargi and Zettlemoyer, Luke and Feichtenhofer, Christoph}, - journal = {arXiv preprint arXiv:2309.16671}, - title = {Demystifying CLIP Data}, - year = 2023} - + title = {Demystifying CLIP Data}, + author = {Xu, Hu and Xie, Saining and Tan, Xiaoqing Ellen and Huang, Po-Yao and Howes, Russell and Sharma, Vasu and Li, Shang-Wen and Ghosh, Gargi and Zettlemoyer, Luke and Feichtenhofer, Christoph}, + year = 2023, + journal = {arXiv preprint arXiv:2309.16671} +} @article{xu2023federated, - author = {Xu, Zheng and Zhang, Yanxiang and Andrew, Galen and Choquette-Choo, Christopher A and Kairouz, Peter and McMahan, H Brendan and Rosenstock, Jesse and Zhang, Yuanbo}, - journal = {arXiv preprint arXiv:2305.18465}, - title = {Federated Learning of Gboard Language Models with Differential Privacy}, - year = 2023} - + title = {Federated Learning of Gboard Language Models with Differential Privacy}, + author = {Xu, Zheng and Zhang, Yanxiang and Andrew, Galen and Choquette-Choo, Christopher A and Kairouz, Peter and McMahan, H Brendan and Rosenstock, Jesse and Zhang, Yuanbo}, + year = 2023, + journal = {arXiv preprint arXiv:2305.18465} +} @article{yamashita2023coffee, - author = {Yamashita, Jo{\~a}o Vitor Yukio Bordin and Leite, Jo{\~a}o Paulo RR}, - journal = {Smart Agricultural Technology}, - pages = 100183, - publisher = {Elsevier}, - title = {Coffee disease classification at the edge using deep learning}, - volume = 4, - year = 2023} - + title = {Coffee disease classification at the edge using deep learning}, + author = {Yamashita, Jo{\~a}o Vitor Yukio Bordin and Leite, Jo{\~a}o Paulo RR}, + year = 2023, + journal = {Smart Agricultural Technology}, + publisher = {Elsevier}, + volume = 4, + pages = 100183 +} @misc{yang2020coexploration, + title = {Co-Exploration of Neural Architectures and Heterogeneous ASIC Accelerator Designs Targeting Multiple Tasks}, + author = {Lei Yang and Zheyu Yan and Meng Li and Hyoukjun Kwon and Liangzhen Lai and Tushar Krishna and Vikas Chandra and Weiwen Jiang and Yiyu Shi}, + year = 2020, archiveprefix = {arXiv}, - author = {Lei Yang and Zheyu Yan and Meng Li and Hyoukjun Kwon and Liangzhen Lai and Tushar Krishna and Vikas Chandra and Weiwen Jiang and Yiyu Shi}, - eprint = {2002.04116}, - primaryclass = {cs.LG}, - title = {Co-Exploration of Neural Architectures and Heterogeneous ASIC Accelerator Designs Targeting Multiple Tasks}, - year = 2020} - + eprint = {2002.04116}, + primaryclass = {cs.LG} +} @inproceedings{yang2023online, - author = {Yang, Tien-Ju and Xiao, Yonghui and Motta, Giovanni and Beaufays, Fran{\c{c}}oise and Mathews, Rajiv and Chen, Mingqing}, - booktitle = {ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)}, - organization = {IEEE}, - pages = {1--5}, - title = {Online Model Compression for Federated Learning with Large Models}, - year = 2023} - + title = {Online Model Compression for Federated Learning with Large Models}, + author = {Yang, Tien-Ju and Xiao, Yonghui and Motta, Giovanni and Beaufays, Fran{\c{c}}oise and Mathews, Rajiv and Chen, Mingqing}, + year = 2023, + booktitle = {ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)}, + pages = {1--5}, + organization = {IEEE} +} +@misc{yik2023neurobench, + title = {NeuroBench: Advancing Neuromorphic Computing through Collaborative, Fair and Representative Benchmarking}, + author = {Jason Yik and Soikat Hasan Ahmed and Zergham Ahmed and Brian Anderson and Andreas G. Andreou and Chiara Bartolozzi and Arindam Basu and Douwe den Blanken and Petrut Bogdan and Sander Bohte and Younes Bouhadjar and Sonia Buckley and Gert Cauwenberghs and Federico Corradi and Guido de Croon and Andreea Danielescu and Anurag Daram and Mike Davies and Yigit Demirag and Jason Eshraghian and Jeremy Forest and Steve Furber and Michael Furlong and Aditya Gilra and Giacomo Indiveri and Siddharth Joshi and Vedant Karia and Lyes Khacef and James C. Knight and Laura Kriener and Rajkumar Kubendran and Dhireesha Kudithipudi and Gregor Lenz and Rajit Manohar and Christian Mayr and Konstantinos Michmizos and Dylan Muir and Emre Neftci and Thomas Nowotny and Fabrizio Ottati and Ayca Ozcelikkale and Noah Pacik-Nelson and Priyadarshini Panda and Sun Pao-Sheng and Melika Payvand and Christian Pehle and Mihai A. Petrovici and Christoph Posch and Alpha Renner and Yulia Sandamirskaya and Clemens JS Schaefer and Andr{\'e} van Schaik and Johannes Schemmel and Catherine Schuman and Jae-sun Seo and Sadique Sheik and Sumit Bam Shrestha and Manolis Sifalakis and Amos Sironi and Kenneth Stewart and Terrence C. Stewart and Philipp Stratmann and Guangzhi Tang and Jonathan Timcheck and Marian Verhelst and Craig M. Vineyard and Bernhard Vogginger and Amirreza Yousefzadeh and Biyan Zhou and Fatima Tuz Zohora and Charlotte Frenkel and Vijay Janapa Reddi}, + year = 2023, + archiveprefix = {arXiv}, + eprint = {2304.04640}, + primaryclass = {cs.AI} +} +@article{young2018recent, + title = {Recent trends in deep learning based natural language processing}, + author = {Young, Tom and Hazarika, Devamanyu and Poria, Soujanya and Cambria, Erik}, + year = 2018, + journal = {ieee Computational intelligenCe magazine}, + publisher = {IEEE}, + volume = 13, + number = 3, + pages = {55--75} +} +@inproceedings{zafrir2019q8bert, + title = {Q8bert: Quantized 8bit bert}, + author = {Zafrir, Ofir and Boudoukh, Guy and Izsak, Peter and Wasserblat, Moshe}, + year = 2019, + booktitle = {2019 Fifth Workshop on Energy Efficient Machine Learning and Cognitive Computing-NeurIPS Edition (EMC2-NIPS)}, + pages = {36--39}, + organization = {IEEE} +} @inproceedings{zennaro2022tinyml, - author = {Zennaro, Marco and Plancher, Brian and Reddi, V Janapa}, - booktitle = {The UN 7th Multi-stakeholder Forum on Science, Technology and Innovation for the Sustainable Development Goals}, - pages = {2022--05}, - title = {TinyML: applied AI for development}, - year = 2022} - + title = {TinyML: applied AI for development}, + author = {Zennaro, Marco and Plancher, Brian and Reddi, V Janapa}, + year = 2022, + booktitle = {The UN 7th Multi-stakeholder Forum on Science, Technology and Innovation for the Sustainable Development Goals}, + pages = {2022--05} +} @article{zennarobridging, - author = {Zennaro, Marco and Plancher, Brian and Reddi, Vijay Janapa}, - title = {Bridging the Digital Divide: the Promising Impact of TinyML for Developing Countries}} - + title = {Bridging the Digital Divide: the Promising Impact of TinyML for Developing Countries}, + author = {Zennaro, Marco and Plancher, Brian and Reddi, Vijay Janapa} +} @inproceedings{Zhang_2020_CVPR_Workshops, - author = {Zhang, Li Lyna and Yang, Yuqing and Jiang, Yuhang and Zhu, Wenwu and Liu, Yunxin}, - booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops}, - month = {June}, - title = {Fast Hardware-Aware Neural Architecture Search}, - year = 2020} - + title = {Fast Hardware-Aware Neural Architecture Search}, + author = {Zhang, Li Lyna and Yang, Yuqing and Jiang, Yuhang and Zhu, Wenwu and Liu, Yunxin}, + year = 2020, + month = {June}, + booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops} +} +@inproceedings{zhang2015fpga, + title = {FPGA-based Accelerator Design for Deep Convolutional Neural Networks Proceedings of the 2015 ACM}, + author = {Zhang, Chen and Li, Peng and Sun, Guangyu and Guan, Yijin and Xiao, Bingjun and Cong, Jason Optimizing}, + year = 2015, + booktitle = {SIGDA International Symposium on Field-Programmable Gate Arrays-FPGA}, + volume = 15, + pages = {161--170} +} +@article{Zhang2017, + title = {Highly wearable cuff-less blood pressure and heart rate monitoring with single-arm electrocardiogram and photoplethysmogram signals}, + author = {Zhang, Qingxue and Zhou, Dian and Zeng, Xuan}, + year = 2017, + month = {Feb}, + day = {06}, + journal = {BioMedical Engineering OnLine}, + volume = 16, + number = 1, + pages = 23, + doi = {10.1186/s12938-017-0317-z}, + issn = {1475-925X}, + url = {https://doi.org/10.1186/s12938-017-0317-z}, + bdsk-url-1 = {https://doi.org/10.1186/s12938-017-0317-z} +} +@article{zhang2018review, + title = {Review on the research and practice of deep learning and reinforcement learning in smart grids}, + author = {Zhang, Dongxia and Han, Xiaoqing and Deng, Chunyu}, + year = 2018, + journal = {CSEE Journal of Power and Energy Systems}, + publisher = {CSEE}, + volume = 4, + number = 3, + pages = {362--370} +} @misc{zhang2019autoshrink, + title = {AutoShrink: A Topology-aware NAS for Discovering Efficient Neural Architecture}, + author = {Tunhou Zhang and Hsin-Pai Cheng and Zhenwen Li and Feng Yan and Chengyu Huang and Hai Li and Yiran Chen}, + year = 2019, archiveprefix = {arXiv}, - author = {Tunhou Zhang and Hsin-Pai Cheng and Zhenwen Li and Feng Yan and Chengyu Huang and Hai Li and Yiran Chen}, - eprint = {1911.09251}, - primaryclass = {cs.LG}, - title = {AutoShrink: A Topology-aware NAS for Discovering Efficient Neural Architecture}, - year = 2019} - + eprint = {1911.09251}, + primaryclass = {cs.LG} +} @article{zhao2018federated, - author = {Zhao, Yue and Li, Meng and Lai, Liangzhen and Suda, Naveen and Civin, Damon and Chandra, Vikas}, - journal = {arXiv preprint arXiv:1806.00582}, - title = {Federated learning with non-iid data}, - year = 2018} - + title = {Federated learning with non-iid data}, + author = {Zhao, Yue and Li, Meng and Lai, Liangzhen and Suda, Naveen and Civin, Damon and Chandra, Vikas}, + year = 2018, + journal = {arXiv preprint arXiv:1806.00582} +} +@inproceedings{zhao2018fpga, + title = {FPGA-based remote power side-channel attacks}, + author = {Zhao, Mark and Suh, G Edward}, + year = 2018, + booktitle = {2018 IEEE Symposium on Security and Privacy (SP)}, + pages = {229--244}, + date-added = {2023-11-22 17:08:21 -0500}, + date-modified = {2023-11-22 17:09:07 -0500}, + organization = {IEEE} +} @misc{zhou_deep_2023, - annote = {Comment: Code is available at https://github.com/zhoudw-zdw/CIL\_Survey/}, - author = {Zhou, Da-Wei and Wang, Qi-Wei and Qi, Zhi-Hong and Ye, Han-Jia and Zhan, De-Chuan and Liu, Ziwei}, - file = {Zhou et al. - 2023 - Deep Class-Incremental Learning A Survey.pdf:/Users/alex/Zotero/storage/859VZG7W/Zhou et al. - 2023 - Deep Class-Incremental Learning A Survey.pdf:application/pdf}, - keywords = {Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning}, - language = {en}, - month = feb, - note = {arXiv:2302.03648 [cs]}, - publisher = {arXiv}, - shorttitle = {Deep {Class}-{Incremental} {Learning}}, - title = {Deep {Class}-{Incremental} {Learning}: {A} {Survey}}, - url = {http://arxiv.org/abs/2302.03648}, - urldate = {2023-10-26}, - year = 2023, - Bdsk-Url-1 = {http://arxiv.org/abs/2302.03648}} - -@misc{noauthor_who_nodate, - title = {Who {Invented} the {Microprocessor}? - {CHM}}, - url = {https://computerhistory.org/blog/who-invented-the-microprocessor/}, - urldate = {2023-11-07}, - Bdsk-Url-1 = {https://computerhistory.org/blog/who-invented-the-microprocessor/}} - -@book{weik_survey_1955, - author = {Weik, Martin H.}, - language = {en}, - publisher = {Ballistic Research Laboratories}, - title = {A {Survey} of {Domestic} {Electronic} {Digital} {Computing} {Systems}}, - year = {1955}} - -@inproceedings{brown_language_2020, - abstract = {We demonstrate that scaling up language models greatly improves task-agnostic, few-shot performance, sometimes even becoming competitive with prior state-of-the-art fine-tuning approaches. Specifically, we train GPT-3, an autoregressive language model with 175 billion parameters, 10x more than any previous non-sparse language model, and test its performance in the few-shot setting. For all tasks, GPT-3 is applied without any gradient updates or fine-tuning, with tasks and few-shot demonstrations specified purely via text interaction with the model. GPT-3 achieves strong performance on many NLP datasets, including translation, question-answering, and cloze tasks. We also identify some datasets where GPT-3's few-shot learning still struggles, as well as some datasets where GPT-3 faces methodological issues related to training on large web corpora.}, - author = {Brown, Tom and Mann, Benjamin and Ryder, Nick and Subbiah, Melanie and Kaplan, Jared D and Dhariwal, Prafulla and Neelakantan, Arvind and Shyam, Pranav and Sastry, Girish and Askell, Amanda and Agarwal, Sandhini and Herbert-Voss, Ariel and Krueger, Gretchen and Henighan, Tom and Child, Rewon and Ramesh, Aditya and Ziegler, Daniel and Wu, Jeffrey and Winter, Clemens and Hesse, Chris and Chen, Mark and Sigler, Eric and Litwin, Mateusz and Gray, Scott and Chess, Benjamin and Clark, Jack and Berner, Christopher and McCandlish, Sam and Radford, Alec and Sutskever, Ilya and Amodei, Dario}, - booktitle = {Advances in {Neural} {Information} {Processing} {Systems}}, - pages = {1877--1901}, - publisher = {Curran Associates, Inc.}, - title = {Language {Models} are {Few}-{Shot} {Learners}}, - url = {https://proceedings.neurips.cc/paper_files/paper/2020/hash/1457c0d6bfcb4967418bfb8ac142f64a-Abstract.html}, - urldate = {2023-11-07}, - volume = {33}, - year = {2020}, - Bdsk-Url-1 = {https://proceedings.neurips.cc/paper_files/paper/2020/hash/1457c0d6bfcb4967418bfb8ac142f64a-Abstract.html}} - -@misc{jia_dissecting_2018, - abstract = {Every year, novel NVIDIA GPU designs are introduced. This rapid architectural and technological progression, coupled with a reluctance by manufacturers to disclose low-level details, makes it difficult for even the most proficient GPU software designers to remain up-to-date with the technological advances at a microarchitectural level. To address this dearth of public, microarchitectural-level information on the novel NVIDIA GPUs, independent researchers have resorted to microbenchmarks-based dissection and discovery. This has led to a prolific line of publications that shed light on instruction encoding, and memory hierarchy's geometry and features at each level. Namely, research that describes the performance and behavior of the Kepler, Maxwell and Pascal architectures. In this technical report, we continue this line of research by presenting the microarchitectural details of the NVIDIA Volta architecture, discovered through microbenchmarks and instruction set disassembly. Additionally, we compare quantitatively our Volta findings against its predecessors, Kepler, Maxwell and Pascal.}, - author = {Jia, Zhe and Maggioni, Marco and Staiger, Benjamin and Scarpazza, Daniele P.}, - keywords = {Computer Science - Distributed, Parallel, and Cluster Computing, Computer Science - Performance}, - month = apr, - note = {arXiv:1804.06826 [cs]}, - publisher = {arXiv}, - title = {Dissecting the {NVIDIA} {Volta} {GPU} {Architecture} via {Microbenchmarking}}, - url = {http://arxiv.org/abs/1804.06826}, - urldate = {2023-11-07}, - year = {2018}, - Bdsk-Url-1 = {http://arxiv.org/abs/1804.06826}} - -@article{jia2019beyond, - author = {Jia, Zhihao and Zaharia, Matei and Aiken, Alex}, - journal = {Proceedings of Machine Learning and Systems}, - pages = {1--13}, - title = {Beyond Data and Model Parallelism for Deep Neural Networks.}, - volume = {1}, - year = {2019}} - -@inproceedings{raina_large-scale_2009, - address = {Montreal Quebec Canada}, - author = {Raina, Rajat and Madhavan, Anand and Ng, Andrew Y.}, - booktitle = {Proceedings of the 26th {Annual} {International} {Conference} on {Machine} {Learning}}, - doi = {10.1145/1553374.1553486}, - isbn = {978-1-60558-516-1}, - language = {en}, - month = jun, - pages = {873--880}, - publisher = {ACM}, - title = {Large-scale deep unsupervised learning using graphics processors}, - url = {https://dl.acm.org/doi/10.1145/1553374.1553486}, - urldate = {2023-11-07}, - year = {2009}, - Bdsk-Url-1 = {https://dl.acm.org/doi/10.1145/1553374.1553486}, - Bdsk-Url-2 = {https://doi.org/10.1145/1553374.1553486}} - -@misc{noauthor_amd_nodate, - title = {{AMD} {Radeon} {RX} 7000 {Series} {Desktop} {Graphics} {Cards}}, - url = {https://www.amd.com/en/graphics/radeon-rx-graphics}, - urldate = {2023-11-07}, - Bdsk-Url-1 = {https://www.amd.com/en/graphics/radeon-rx-graphics}} - -@misc{noauthor_intel_nodate, - abstract = {Find out how Intel{\textregistered} Arc Graphics unlock lifelike gaming and seamless content creation.}, - journal = {Intel}, - language = {en}, - title = {Intel{\textregistered} {Arc}{\texttrademark} {Graphics} {Overview}}, - url = {https://www.intel.com/content/www/us/en/products/details/discrete-gpus/arc.html}, - urldate = {2023-11-07}, - Bdsk-Url-1 = {https://www.intel.com/content/www/us/en/products/details/discrete-gpus/arc.html}} - -@article{lindholm_nvidia_2008, - abstract = {To enable flexible, programmable graphics and high-performance computing, NVIDIA has developed the Tesla scalable unified graphics and parallel computing architecture. Its scalable parallel array of processors is massively multithreaded and programmable in C or via graphics APIs.}, - author = {Lindholm, Erik and Nickolls, John and Oberman, Stuart and Montrym, John}, - doi = {10.1109/MM.2008.31}, - issn = {1937-4143}, - journal = {IEEE Micro}, - month = mar, - note = {Conference Name: IEEE Micro}, - number = {2}, - pages = {39--55}, - shorttitle = {{NVIDIA} {Tesla}}, - title = {{NVIDIA} {Tesla}: {A} {Unified} {Graphics} and {Computing} {Architecture}}, - url = {https://ieeexplore.ieee.org/document/4523358}, - urldate = {2023-11-07}, - volume = {28}, - year = {2008}, - Bdsk-Url-1 = {https://ieeexplore.ieee.org/document/4523358}, - Bdsk-Url-2 = {https://doi.org/10.1109/MM.2008.31}} - -@article{dally_evolution_2021, - abstract = {Graphics processing units (GPUs) power today's fastest supercomputers, are the dominant platform for deep learning, and provide the intelligence for devices ranging from self-driving cars to robots and smart cameras. They also generate compelling photorealistic images at real-time frame rates. GPUs have evolved by adding features to support new use cases. NVIDIA's GeForce 256, the first GPU, was a dedicated processor for real-time graphics, an application that demands large amounts of floating-point arithmetic for vertex and fragment shading computations and high memory bandwidth. As real-time graphics advanced, GPUs became programmable. The combination of programmability and floating-point performance made GPUs attractive for running scientific applications. Scientists found ways to use early programmable GPUs by casting their calculations as vertex and fragment shaders. GPUs evolved to meet the needs of scientific users by adding hardware for simpler programming, double-precision floating-point arithmetic, and resilience.}, - author = {Dally, William J. and Keckler, Stephen W. and Kirk, David B.}, - doi = {10.1109/MM.2021.3113475}, - issn = {1937-4143}, - journal = {IEEE Micro}, - month = nov, - note = {Conference Name: IEEE Micro}, - number = {6}, - pages = {42--51}, - title = {Evolution of the {Graphics} {Processing} {Unit} ({GPU})}, - url = {https://ieeexplore.ieee.org/document/9623445}, - urldate = {2023-11-07}, - volume = {41}, - year = {2021}, - Bdsk-Url-1 = {https://ieeexplore.ieee.org/document/9623445}, - Bdsk-Url-2 = {https://doi.org/10.1109/MM.2021.3113475}} - -@article{demler_ceva_2020, - author = {Demler, Mike}, - language = {en}, - title = {{CEVA} {SENSPRO} {FUSES} {AI} {AND} {VECTOR} {DSP}}, - year = {2020}} - -@misc{noauthor_google_2023, - abstract = {Tensor G3 on Pixel 8 and Pixel 8 Pro is more helpful, more efficient and more powerful.}, - journal = {Google}, - language = {en-us}, - month = oct, - shorttitle = {Google {Tensor} {G3}}, - title = {Google {Tensor} {G3}: {The} new chip that gives your {Pixel} an {AI} upgrade}, - url = {https://blog.google/products/pixel/google-tensor-g3-pixel-8/}, - urldate = {2023-11-07}, - year = {2023}, - Bdsk-Url-1 = {https://blog.google/products/pixel/google-tensor-g3-pixel-8/}} - -@misc{noauthor_hexagon_nodate, - abstract = {The Hexagon DSP processor has both CPU and DSP functionality to support deeply embedded processing needs of the mobile platform for both multimedia and modem functions.}, - journal = {Qualcomm Developer Network}, - language = {en}, - title = {Hexagon {DSP} {SDK} {Processor}}, - url = {https://developer.qualcomm.com/software/hexagon-dsp-sdk/dsp-processor}, - urldate = {2023-11-07}, - Bdsk-Url-1 = {https://developer.qualcomm.com/software/hexagon-dsp-sdk/dsp-processor}} - -@misc{noauthor_evolution_2023, - abstract = {To complement the extensive perspective of another Market Update feature article on DSP Products and Applications, published in the November 2020 edition, audioXpress was honored to have the valuable contribution from one of the main suppliers in the field. In this article, Youval Nachum, CEVA's Senior Product Marketing Manager, writes about \"The Evolution of Audio DSPs,\" discussing how DSP technology has evolved, its impact on the user experience, and what the future of DSP has in store for us.}, - journal = {audioXpress}, - language = {en}, - month = oct, - title = {The {Evolution} of {Audio} {DSPs}}, - url = {https://audioxpress.com/article/the-evolution-of-audio-dsps}, - urldate = {2023-11-07}, - year = {2023}, - Bdsk-Url-1 = {https://audioxpress.com/article/the-evolution-of-audio-dsps}} - -@article{xiong_mri-based_2021, - abstract = {Brain tumor segmentation is a challenging problem in medical image processing and analysis. It is a very time-consuming and error-prone task. In order to reduce the burden on physicians and improve the segmentation accuracy, the computer-aided detection (CAD) systems need to be developed. Due to the powerful feature learning ability of the deep learning technology, many deep learning-based methods have been applied to the brain tumor segmentation CAD systems and achieved satisfactory accuracy. However, deep learning neural networks have high computational complexity, and the brain tumor segmentation process consumes significant time. Therefore, in order to achieve the high segmentation accuracy of brain tumors and obtain the segmentation results efficiently, it is very demanding to speed up the segmentation process of brain tumors.}, - author = {Xiong, Siyu and Wu, Guoqing and Fan, Xitian and Feng, Xuan and Huang, Zhongcheng and Cao, Wei and Zhou, Xuegong and Ding, Shijin and Yu, Jinhua and Wang, Lingli and Shi, Zhifeng}, - doi = {10.1186/s12859-021-04347-6}, - issn = {1471-2105}, - journal = {BMC Bioinformatics}, - keywords = {Brain tumor segmatation, FPGA acceleration, Neural network}, - month = sep, - number = {1}, - pages = {421}, - title = {{MRI}-based brain tumor segmentation using {FPGA}-accelerated neural network}, - url = {https://doi.org/10.1186/s12859-021-04347-6}, - urldate = {2023-11-07}, - volume = {22}, - year = {2021}, - Bdsk-Url-1 = {https://doi.org/10.1186/s12859-021-04347-6}} - -@article{gwennap_certus-nx_nodate, - author = {Gwennap, Linley}, - language = {en}, - title = {Certus-{NX} {Innovates} {General}-{Purpose} {FPGAs}}} - -@misc{noauthor_fpga_nodate, - title = {{FPGA} {Architecture} {Overview}}, - url = {https://www.intel.com/content/www/us/en/docs/oneapi-fpga-add-on/optimization-guide/2023-1/fpga-architecture-overview.html}, - urldate = {2023-11-07}, - Bdsk-Url-1 = {https://www.intel.com/content/www/us/en/docs/oneapi-fpga-add-on/optimization-guide/2023-1/fpga-architecture-overview.html}} - -@misc{noauthor_what_nodate, - abstract = {What is an FPGA - Field Programmable Gate Arrays are semiconductor devices that are based around a matrix of configurable logic blocks (CLBs) connected via programmable interconnects. FPGAs can be reprogrammed to desired application or functionality requirements after manufacturing.}, - journal = {AMD}, - language = {en}, - shorttitle = {What is an {FPGA}?}, - title = {What is an {FPGA}? {Field} {Programmable} {Gate} {Array}}, - url = {https://www.xilinx.com/products/silicon-devices/fpga/what-is-an-fpga.html}, - urldate = {2023-11-07}, - Bdsk-Url-1 = {https://www.xilinx.com/products/silicon-devices/fpga/what-is-an-fpga.html}} - -@article{putnam_reconfigurable_2014, - abstract = {Datacenter workloads demand high computational capabilities, flexibility, power efficiency, and low cost. It is challenging to improve all of these factors simultaneously. To advance datacenter capabilities beyond what commodity server designs can provide, we have designed and built a composable, reconfigurablefabric to accelerate portions of large-scale software services. Each instantiation of the fabric consists of a 6x8 2-D torus of high-end Stratix V FPGAs embedded into a half-rack of 48 machines. One FPGA is placed into each server, accessible through PCIe, and wired directly to other FPGAs with pairs of 10 Gb SAS cables - In this paper, we describe a medium-scale deployment of this fabric on a bed of 1,632 servers, and measure its efficacy in accelerating the Bing web search engine. We describe the requirements and architecture of the system, detail the critical engineering challenges and solutions needed to make the system robust in the presence of failures, and measure the performance, power, and resilience of the system when ranking candidate documents. Under high load, the largescale reconfigurable fabric improves the ranking throughput of each server by a factor of 95\% for a fixed latency distribution--- or, while maintaining equivalent throughput, reduces the tail latency by 29\%}, - author = {Putnam, Andrew and Caulfield, Adrian M. and Chung, Eric S. and Chiou, Derek and Constantinides, Kypros and Demme, John and Esmaeilzadeh, Hadi and Fowers, Jeremy and Gopal, Gopi Prashanth and Gray, Jan and Haselman, Michael and Hauck, Scott and Heil, Stephen and Hormati, Amir and Kim, Joo-Young and Lanka, Sitaram and Larus, James and Peterson, Eric and Pope, Simon and Smith, Aaron and Thong, Jason and Xiao, Phillip Yi and Burger, Doug}, - doi = {10.1145/2678373.2665678}, - issn = {0163-5964}, - journal = {ACM SIGARCH Computer Architecture News}, - language = {en}, - month = oct, - number = {3}, - pages = {13--24}, - title = {A reconfigurable fabric for accelerating large-scale datacenter services}, - url = {https://dl.acm.org/doi/10.1145/2678373.2665678}, - urldate = {2023-11-07}, - volume = {42}, - year = {2014}, - Bdsk-Url-1 = {https://dl.acm.org/doi/10.1145/2678373.2665678}, - Bdsk-Url-2 = {https://doi.org/10.1145/2678373.2665678}} - -@misc{noauthor_project_nodate, - title = {Project {Catapult} - {Microsoft} {Research}}, - url = {https://www.microsoft.com/en-us/research/project/project-catapult/}, - urldate = {2023-11-07}, - Bdsk-Url-1 = {https://www.microsoft.com/en-us/research/project/project-catapult/}} - -@misc{dean_jeff_numbers_nodate, - author = {Dean. Jeff}, - title = {Numbers {Everyone} {Should} {Know}}, - url = {https://brenocon.com/dean_perf.html}, - urldate = {2023-11-07}, - Bdsk-Url-1 = {https://brenocon.com/dean_perf.html}} - -@misc{bailey_enabling_2018, - abstract = {Enabling Cheaper Design, At what point does cheaper design enable a significant growth in custom semiconductor content? Not everyone is onboard with the idea.}, - author = {Bailey, Brian}, - journal = {Semiconductor Engineering}, - language = {en-US}, - month = sep, - title = {Enabling {Cheaper} {Design}}, - url = {https://semiengineering.com/enabling-cheaper-design/}, - urldate = {2023-11-07}, - year = {2018}, - Bdsk-Url-1 = {https://semiengineering.com/enabling-cheaper-design/}} - -@misc{noauthor_integrated_2023, - abstract = {An integrated circuit (also known as an IC, a chip, or a microchip) is a set of electronic circuits on one small flat piece of semiconductor material, usually silicon. Large numbers of miniaturized transistors and other electronic components are integrated together on the chip. This results in circuits that are orders of magnitude smaller, faster, and less expensive than those constructed of discrete components, allowing a large transistor count. -The IC's mass production capability, reliability, and building-block approach to integrated circuit design have ensured the rapid adoption of standardized ICs in place of designs using discrete transistors. ICs are now used in virtually all electronic equipment and have revolutionized the world of electronics. Computers, mobile phones and other home appliances are now essential parts of the structure of modern societies, made possible by the small size and low cost of ICs such as modern computer processors and microcontrollers. -Very-large-scale integration was made practical by technological advancements in semiconductor device fabrication. Since their origins in the 1960s, the size, speed, and capacity of chips have progressed enormously, driven by technical advances that fit more and more transistors on chips of the same size -- a modern chip may have many billions of transistors in an area the size of a human fingernail. These advances, roughly following Moore's law, make the computer chips of today possess millions of times the capacity and thousands of times the speed of the computer chips of the early 1970s. -ICs have three main advantages over discrete circuits: size, cost and performance. The size and cost is low because the chips, with all their components, are printed as a unit by photolithography rather than being constructed one transistor at a time. Furthermore, packaged ICs use much less material than discrete circuits. Performance is high because the IC's components switch quickly and consume comparatively little power because of their small size and proximity. The main disadvantage of ICs is the high initial cost of designing them and the enormous capital cost of factory construction. This high initial cost means ICs are only commercially viable when high production volumes are anticipated.}, - copyright = {Creative Commons Attribution-ShareAlike License}, - journal = {Wikipedia}, - language = {en}, - month = nov, - note = {Page Version ID: 1183537457}, - title = {Integrated circuit}, - url = {https://en.wikipedia.org/w/index.php?title=Integrated_circuit&oldid=1183537457}, - urldate = {2023-11-07}, - year = {2023}, - Bdsk-Url-1 = {https://en.wikipedia.org/w/index.php?title=Integrated_circuit&oldid=1183537457}} - -@article{el-rayis_reconfigurable_nodate, - author = {El-Rayis, Ahmed Osman}, - language = {en}, - title = {Reconfigurable {Architectures} for the {Next} {Generation} of {Mobile} {Device} {Telecommunications} {Systems}}} - -@misc{noauthor_intel_nodate, - abstract = {View Intel{\textregistered} Stratix{\textregistered} 10 NX FPGAs and find product specifications, features, applications and more.}, - journal = {Intel}, - language = {en}, - title = {Intel{\textregistered} {Stratix}{\textregistered} 10 {NX} {FPGA} {Overview} - {High} {Performance} {Stratix}{\textregistered} {FPGA}}, - url = {https://www.intel.com/content/www/us/en/products/details/fpga/stratix/10/nx.html}, - urldate = {2023-11-07}, - Bdsk-Url-1 = {https://www.intel.com/content/www/us/en/products/details/fpga/stratix/10/nx.html}} - -@book{patterson2016computer, - author = {Patterson, David A and Hennessy, John L}, - publisher = {Morgan kaufmann}, - title = {Computer organization and design ARM edition: the hardware software interface}, - year = {2016}} - -@article{xiu2019time, - author = {Xiu, Liming}, - journal = {IEEE Solid-State Circuits Magazine}, - number = {1}, - pages = {39--55}, - publisher = {IEEE}, - title = {Time Moore: Exploiting Moore's Law from the perspective of time}, - volume = {11}, - year = {2019}} - -@article{brown2020language, - author = {Brown, Tom and Mann, Benjamin and Ryder, Nick and Subbiah, Melanie and Kaplan, Jared D and Dhariwal, Prafulla and Neelakantan, Arvind and Shyam, Pranav and Sastry, Girish and Askell, Amanda and others}, - journal = {Advances in neural information processing systems}, - pages = {1877--1901}, - title = {Language models are few-shot learners}, - volume = {33}, - year = {2020}} - -@article{cheng2017survey, - author = {Cheng, Yu and Wang, Duo and Zhou, Pan and Zhang, Tao}, - journal = {arXiv preprint arXiv:1710.09282}, - title = {A survey of model compression and acceleration for deep neural networks}, - year = {2017}} - -@article{sze2017efficient, - author = {Sze, Vivienne and Chen, Yu-Hsin and Yang, Tien-Ju and Emer, Joel S}, - journal = {Proceedings of the IEEE}, - number = {12}, - pages = {2295--2329}, - publisher = {Ieee}, - title = {Efficient processing of deep neural networks: A tutorial and survey}, - volume = {105}, - year = {2017}} - -@article{young2018recent, - author = {Young, Tom and Hazarika, Devamanyu and Poria, Soujanya and Cambria, Erik}, - journal = {ieee Computational intelligenCe magazine}, - number = {3}, - pages = {55--75}, - publisher = {IEEE}, - title = {Recent trends in deep learning based natural language processing}, - volume = {13}, - year = {2018}} - -@inproceedings{jacob2018quantization, - author = {Jacob, Benoit and Kligys, Skirmantas and Chen, Bo and Zhu, Menglong and Tang, Matthew and Howard, Andrew and Adam, Hartwig and Kalenichenko, Dmitry}, - booktitle = {Proceedings of the IEEE conference on computer vision and pattern recognition}, - pages = {2704--2713}, - title = {Quantization and training of neural networks for efficient integer-arithmetic-only inference}, - year = {2018}} - -@article{gale2019state, - author = {Gale, Trevor and Elsen, Erich and Hooker, Sara}, - journal = {arXiv preprint arXiv:1902.09574}, - title = {The state of sparsity in deep neural networks}, - year = {2019}} - -@inproceedings{zhang2015fpga, - author = {Zhang, Chen and Li, Peng and Sun, Guangyu and Guan, Yijin and Xiao, Bingjun and Cong, Jason Optimizing}, - booktitle = {SIGDA International Symposium on Field-Programmable Gate Arrays-FPGA}, - pages = {161--170}, - title = {FPGA-based Accelerator Design for Deep Convolutional Neural Networks Proceedings of the 2015 ACM}, - volume = {15}, - year = {2015}} - -@inproceedings{suda2016throughput, - author = {Suda, Naveen and Chandra, Vikas and Dasika, Ganesh and Mohanty, Abinash and Ma, Yufei and Vrudhula, Sarma and Seo, Jae-sun and Cao, Yu}, - booktitle = {Proceedings of the 2016 ACM/SIGDA international symposium on field-programmable gate arrays}, - pages = {16--25}, - title = {Throughput-optimized OpenCL-based FPGA accelerator for large-scale convolutional neural networks}, - year = {2016}} - -@inproceedings{fowers2018configurable, - author = {Fowers, Jeremy and Ovtcharov, Kalin and Papamichael, Michael and Massengill, Todd and Liu, Ming and Lo, Daniel and Alkalay, Shlomi and Haselman, Michael and Adams, Logan and Ghandi, Mahdi and others}, - booktitle = {2018 ACM/IEEE 45th Annual International Symposium on Computer Architecture (ISCA)}, - organization = {IEEE}, - pages = {1--14}, - title = {A configurable cloud-scale DNN processor for real-time AI}, - year = {2018}} - -@article{jia2019beyond, - author = {Jia, Zhihao and Zaharia, Matei and Aiken, Alex}, - journal = {Proceedings of Machine Learning and Systems}, - pages = {1--13}, - title = {Beyond Data and Model Parallelism for Deep Neural Networks.}, - volume = {1}, - year = {2019}} - -@inproceedings{zhu2018benchmarking, - author = {Zhu, Hongyu and Akrout, Mohamed and Zheng, Bojian and Pelegris, Andrew and Jayarajan, Anand and Phanishayee, Amar and Schroeder, Bianca and Pekhimenko, Gennady}, - booktitle = {2018 IEEE International Symposium on Workload Characterization (IISWC)}, - organization = {IEEE}, - pages = {88--100}, - title = {Benchmarking and analyzing deep neural network training}, - year = {2018}} - -@article{samajdar2018scale, - author = {Samajdar, Ananda and Zhu, Yuhao and Whatmough, Paul and Mattina, Matthew and Krishna, Tushar}, - journal = {arXiv preprint arXiv:1811.02883}, - title = {Scale-sim: Systolic cnn accelerator simulator}, - year = {2018}} - -@inproceedings{munshi2009opencl, - author = {Munshi, Aaftab}, - booktitle = {2009 IEEE Hot Chips 21 Symposium (HCS)}, - doi = {10.1109/HOTCHIPS.2009.7478342}, - pages = {1-314}, - title = {The OpenCL specification}, - year = {2009}, - Bdsk-Url-1 = {https://doi.org/10.1109/HOTCHIPS.2009.7478342}} - -@inproceedings{luebke2008cuda, - author = {Luebke, David}, - booktitle = {2008 5th IEEE International Symposium on Biomedical Imaging: From Nano to Macro}, - doi = {10.1109/ISBI.2008.4541126}, - pages = {836-838}, - title = {CUDA: Scalable parallel programming for high-performance scientific computing}, - year = {2008}, - Bdsk-Url-1 = {https://doi.org/10.1109/ISBI.2008.4541126}} - -@misc{segal1999opengl, - author = {Segal, Mark and Akeley, Kurt}, - title = {The OpenGL graphics system: A specification (version 1.1)}, - year = {1999}} - -@inproceedings{gannot1994verilog, - author = {Gannot, G. and Ligthart, M.}, - booktitle = {International Verilog HDL Conference}, - doi = {10.1109/IVC.1994.323743}, - pages = {86-92}, - title = {Verilog HDL based FPGA design}, - year = {1994}, - Bdsk-Url-1 = {https://doi.org/10.1109/IVC.1994.323743}} - -@article{binkert2011gem5, - author = {Binkert, Nathan and Beckmann, Bradford and Black, Gabriel and Reinhardt, Steven K and Saidi, Ali and Basu, Arkaprava and Hestness, Joel and Hower, Derek R and Krishna, Tushar and Sardashti, Somayeh and others}, - journal = {ACM SIGARCH computer architecture news}, - number = {2}, - pages = {1--7}, - publisher = {ACM New York, NY, USA}, - title = {The gem5 simulator}, - volume = {39}, - year = {2011}} - -@article{Vivet2021, - author = {Vivet, Pascal and Guthmuller, Eric and Thonnart, Yvain and Pillonnet, Gael and Fuguet, C{\'e}sar and Miro-Panades, Ivan and Moritz, Guillaume and Durupt, Jean and Bernard, Christian and Varreau, Didier and Pontes, Julian and Thuries, S{\'e}bastien and Coriat, David and Harrand, Michel and Dutoit, Denis and Lattard, Didier and Arnaud, Lucile and Charbonnier, Jean and Coudrain, Perceval and Garnier, Arnaud and Berger, Fr{\'e}d{\'e}ric and Gueugnot, Alain and Greiner, Alain and Meunier, Quentin L. and Farcy, Alexis and Arriordaz, Alexandre and Ch{\'e}ramy, S{\'e}verine and Clermidy, Fabien}, - doi = {10.1109/JSSC.2020.3036341}, - journal = {IEEE Journal of Solid-State Circuits}, - number = {1}, - pages = {79-97}, - title = {IntAct: A 96-Core Processor With Six Chiplets 3D-Stacked on an Active Interposer With Distributed Interconnects and Integrated Power Management}, - volume = {56}, - year = {2021}, - Bdsk-Url-1 = {https://doi.org/10.1109/JSSC.2020.3036341}} - -@article{schuman2022, - author = {Schuman, Catherine D and Kulkarni, Shruti R and Parsa, Maryam and Mitchell, J Parker and Date, Prasanna and Kay, Bill}, - journal = {Nature Computational Science}, - number = {1}, - pages = {10--19}, - publisher = {Nature Publishing Group US New York}, - title = {Opportunities for neuromorphic computing algorithms and applications}, - volume = {2}, - year = {2022}} - -@article{markovic2020, - author = {Markovi{\'c}, Danijela and Mizrahi, Alice and Querlioz, Damien and Grollier, Julie}, - journal = {Nature Reviews Physics}, - number = {9}, - pages = {499--510}, - publisher = {Nature Publishing Group UK London}, - title = {Physics for neuromorphic computing}, - volume = {2}, - year = {2020}} - -@article{furber2016large, - author = {Furber, Steve}, - journal = {Journal of neural engineering}, - number = {5}, - pages = {051001}, - publisher = {IOP Publishing}, - title = {Large-scale neuromorphic computing systems}, - volume = {13}, - year = {2016}} - -@article{davies2018loihi, - author = {Davies, Mike and Srinivasa, Narayan and Lin, Tsung-Han and Chinya, Gautham and Cao, Yongqiang and Choday, Sri Harsha and Dimou, Georgios and Joshi, Prasad and Imam, Nabil and Jain, Shweta and others}, - journal = {Ieee Micro}, - number = {1}, - pages = {82--99}, - publisher = {IEEE}, - title = {Loihi: A neuromorphic manycore processor with on-chip learning}, - volume = {38}, - year = {2018}} - -@article{davies2021advancing, - author = {Davies, Mike and Wild, Andreas and Orchard, Garrick and Sandamirskaya, Yulia and Guerra, Gabriel A Fonseca and Joshi, Prasad and Plank, Philipp and Risbud, Sumedh R}, - journal = {Proceedings of the IEEE}, - number = {5}, - pages = {911--934}, - publisher = {IEEE}, - title = {Advancing neuromorphic computing with loihi: A survey of results and outlook}, - volume = {109}, - year = {2021}} - -@article{modha2023neural, - author = {Modha, Dharmendra S and Akopyan, Filipp and Andreopoulos, Alexander and Appuswamy, Rathinakumar and Arthur, John V and Cassidy, Andrew S and Datta, Pallab and DeBole, Michael V and Esser, Steven K and Otero, Carlos Ortega and others}, - journal = {Science}, - number = {6668}, - pages = {329--335}, - publisher = {American Association for the Advancement of Science}, - title = {Neural inference at the frontier of energy, space, and time}, - volume = {382}, - year = {2023}} - -@article{maass1997networks, - author = {Maass, Wolfgang}, - journal = {Neural networks}, - number = {9}, - pages = {1659--1671}, - publisher = {Elsevier}, - title = {Networks of spiking neurons: the third generation of neural network models}, - volume = {10}, - year = {1997}} - -@article{10242251, - author = {Eshraghian, Jason K. and Ward, Max and Neftci, Emre O. and Wang, Xinxin and Lenz, Gregor and Dwivedi, Girish and Bennamoun, Mohammed and Jeong, Doo Seok and Lu, Wei D.}, - doi = {10.1109/JPROC.2023.3308088}, - journal = {Proceedings of the IEEE}, - number = {9}, - pages = {1016-1054}, - title = {Training Spiking Neural Networks Using Lessons From Deep Learning}, - volume = {111}, - year = {2023}, - Bdsk-Url-1 = {https://doi.org/10.1109/JPROC.2023.3308088}} - -@article{chua1971memristor, - author = {Chua, Leon}, - journal = {IEEE Transactions on circuit theory}, - number = {5}, - pages = {507--519}, - publisher = {IEEE}, - title = {Memristor-the missing circuit element}, - volume = {18}, - year = {1971}} - -@article{shastri2021photonics, - author = {Shastri, Bhavin J and Tait, Alexander N and Ferreira de Lima, Thomas and Pernice, Wolfram HP and Bhaskaran, Harish and Wright, C David and Prucnal, Paul R}, - journal = {Nature Photonics}, - number = {2}, - pages = {102--114}, - publisher = {Nature Publishing Group UK London}, - title = {Photonics for artificial intelligence and neuromorphic computing}, - volume = {15}, - year = {2021}} - -@article{haensch2018next, - author = {Haensch, Wilfried and Gokmen, Tayfun and Puri, Ruchir}, - journal = {Proceedings of the IEEE}, - number = {1}, - pages = {108--122}, - publisher = {IEEE}, - title = {The next generation of deep learning hardware: Analog computing}, - volume = {107}, - year = {2018}} - -@article{hazan2021neuromorphic, - author = {Hazan, Avi and Ezra Tsur, Elishai}, - journal = {Frontiers in Neuroscience}, - pages = {627221}, - publisher = {Frontiers Media SA}, - title = {Neuromorphic analog implementation of neural engineering framework-inspired spiking neuron for high-dimensional representation}, - volume = {15}, - year = {2021}} - -@article{gates2009flexible, - author = {Gates, Byron D}, - journal = {Science}, - number = {5921}, - pages = {1566--1567}, - publisher = {American Association for the Advancement of Science}, - title = {Flexible electronics}, - volume = {323}, - year = {2009}} - -@article{musk2019integrated, - author = {Musk, Elon and others}, - journal = {Journal of medical Internet research}, - number = {10}, - pages = {e16194}, - publisher = {JMIR Publications Inc., Toronto, Canada}, - title = {An integrated brain-machine interface platform with thousands of channels}, - volume = {21}, - year = {2019}} - -@article{tang2023flexible, - author = {Tang, Xin and Shen, Hao and Zhao, Siyuan and Li, Na and Liu, Jia}, - journal = {Nature Electronics}, - number = {2}, - pages = {109--118}, - publisher = {Nature Publishing Group UK London}, - title = {Flexible brain--computer interfaces}, - volume = {6}, - year = {2023}} - -@article{tang2022soft, - author = {Tang, Xin and He, Yichun and Liu, Jia}, - journal = {Biophysics Reviews}, - number = {1}, - publisher = {AIP Publishing}, - title = {Soft bioelectronics for cardiac interfaces}, - volume = {3}, - year = {2022}} - -@article{kwon2022flexible, - author = {Kwon, Sun Hwa and Dong, Lin}, - journal = {Nano Energy}, - pages = {107632}, - publisher = {Elsevier}, - title = {Flexible sensors and machine learning for heart monitoring}, - year = {2022}} - -@article{huang2010pseudo, - author = {Huang, Tsung-Ching and Fukuda, Kenjiro and Lo, Chun-Ming and Yeh, Yung-Hui and Sekitani, Tsuyoshi and Someya, Takao and Cheng, Kwang-Ting}, - journal = {IEEE Transactions on Electron Devices}, - number = {1}, - pages = {141--150}, - publisher = {IEEE}, - title = {Pseudo-CMOS: A design style for low-cost and robust flexible electronics}, - volume = {58}, - year = {2010}} - -@article{biggs2021natively, - author = {Biggs, John and Myers, James and Kufel, Jedrzej and Ozer, Emre and Craske, Simon and Sou, Antony and Ramsdale, Catherine and Williamson, Ken and Price, Richard and White, Scott}, - journal = {Nature}, - number = {7868}, - pages = {532--536}, - publisher = {Nature Publishing Group UK London}, - title = {A natively flexible 32-bit Arm microprocessor}, - volume = {595}, - year = {2021}} - -@article{farah2005neuroethics, - author = {Farah, Martha J}, - journal = {Trends in cognitive sciences}, - number = {1}, - pages = {34--40}, - publisher = {Elsevier}, - title = {Neuroethics: the practical and the philosophical}, - volume = {9}, - year = {2005}} - -@article{segura2018ethical, - author = {Segura Anaya, LH and Alsadoon, Abeer and Costadopoulos, Nectar and Prasad, PWC}, - journal = {Science and engineering ethics}, - pages = {1--28}, - publisher = {Springer}, - title = {Ethical implications of user perceptions of wearable devices}, - volume = {24}, - year = {2018}} - -@article{goodyear2017social, - author = {Goodyear, Victoria A}, - journal = {Qualitative research in sport, exercise and health}, - number = {3}, - pages = {285--302}, - publisher = {Taylor \& Francis}, - title = {Social media, apps and wearable technologies: navigating ethical dilemmas and procedures}, - volume = {9}, - year = {2017}} - -@article{roskies2002neuroethics, - author = {Roskies, Adina}, - journal = {Neuron}, - number = {1}, - pages = {21--23}, - publisher = {Elsevier}, - title = {Neuroethics for the new millenium}, - volume = {35}, - year = {2002}} - -@article{duarte2022fastml, - author = {Duarte, Javier and Tran, Nhan and Hawks, Ben and Herwig, Christian and Muhizi, Jules and Prakash, Shvetank and Reddi, Vijay Janapa}, - journal = {arXiv preprint arXiv:2207.07958}, - title = {FastML Science Benchmarks: Accelerating Real-Time Scientific Edge Machine Learning}, - year = {2022}} - -@article{verma2019memory, - author = {Verma, Naveen and Jia, Hongyang and Valavi, Hossein and Tang, Yinqi and Ozatay, Murat and Chen, Lung-Yen and Zhang, Bonan and Deaville, Peter}, - journal = {IEEE Solid-State Circuits Magazine}, - number = {3}, - pages = {43--55}, - publisher = {IEEE}, - title = {In-memory computing: Advances and prospects}, - volume = {11}, - year = {2019}} - -@article{chi2016prime, - author = {Chi, Ping and Li, Shuangchen and Xu, Cong and Zhang, Tao and Zhao, Jishen and Liu, Yongpan and Wang, Yu and Xie, Yuan}, - journal = {ACM SIGARCH Computer Architecture News}, - number = {3}, - pages = {27--39}, - publisher = {ACM New York, NY, USA}, - title = {Prime: A novel processing-in-memory architecture for neural network computation in reram-based main memory}, - volume = {44}, - year = {2016}} - -@article{burr2016recent, - author = {Burr, Geoffrey W and Brightsky, Matthew J and Sebastian, Abu and Cheng, Huai-Yu and Wu, Jau-Yi and Kim, Sangbum and Sosa, Norma E and Papandreou, Nikolaos and Lung, Hsiang-Lan and Pozidis, Haralampos and others}, - journal = {IEEE Journal on Emerging and Selected Topics in Circuits and Systems}, - number = {2}, - pages = {146--162}, - publisher = {IEEE}, - title = {Recent progress in phase-change memory technology}, - volume = {6}, - year = {2016}} - -@article{loh20083d, - author = {Loh, Gabriel H}, - journal = {ACM SIGARCH computer architecture news}, - number = {3}, - pages = {453--464}, - publisher = {ACM New York, NY, USA}, - title = {3D-stacked memory architectures for multi-core processors}, - volume = {36}, - year = {2008}} - -@article{mittal2021survey, - author = {Mittal, Sparsh and Verma, Gaurav and Kaushik, Brajesh and Khanday, Farooq A}, - journal = {Journal of Systems Architecture}, - pages = {102276}, - publisher = {Elsevier}, - title = {A survey of SRAM-based in-memory computing techniques and applications}, - volume = {119}, - year = {2021}} - -@article{wong2012metal, - author = {Wong, H-S Philip and Lee, Heng-Yuan and Yu, Shimeng and Chen, Yu-Sheng and Wu, Yi and Chen, Pang-Shiu and Lee, Byoungil and Chen, Frederick T and Tsai, Ming-Jinn}, - journal = {Proceedings of the IEEE}, - number = {6}, - pages = {1951--1970}, - publisher = {IEEE}, - title = {Metal--oxide RRAM}, - volume = {100}, - year = {2012}} - -@inproceedings{imani2016resistive, - author = {Imani, Mohsen and Rahimi, Abbas and Rosing, Tajana S}, - booktitle = {2016 Design, Automation \& Test in Europe Conference \& Exhibition (DATE)}, - organization = {IEEE}, - pages = {1327--1332}, - title = {Resistive configurable associative memory for approximate computing}, - year = {2016}} - -@article{miller2000optical, - author = {Miller, David AB}, - journal = {IEEE Journal of Selected Topics in Quantum Electronics}, - number = {6}, - pages = {1312--1317}, - publisher = {IEEE}, - title = {Optical interconnects to silicon}, - volume = {6}, - year = {2000}} - -@article{zhou2022photonic, - author = {Zhou, Hailong and Dong, Jianji and Cheng, Junwei and Dong, Wenchan and Huang, Chaoran and Shen, Yichen and Zhang, Qiming and Gu, Min and Qian, Chao and Chen, Hongsheng and others}, - journal = {Light: Science \& Applications}, - number = {1}, - pages = {30}, - publisher = {Nature Publishing Group UK London}, - title = {Photonic matrix multiplication lights up photonic accelerator and beyond}, - volume = {11}, - year = {2022}} - -@article{bains2020business, - author = {Bains, Sunny}, - journal = {Nat. Electron}, - number = {7}, - pages = {348--351}, - title = {The business of building brains}, - volume = {3}, - year = {2020}} - -@article{Hennessy2019-je, - abstract = {Innovations like domain-specific hardware, enhanced security, - open instruction sets, and agile chip development will lead the - way.}, - author = {Hennessy, John L and Patterson, David A}, - copyright = {http://www.acm.org/publications/policies/copyright\_policy\#Background}, - journal = {Commun. ACM}, - language = {en}, - month = jan, - number = 2, - pages = {48--60}, - publisher = {Association for Computing Machinery (ACM)}, - title = {A new golden age for computer architecture}, - volume = 62, - year = 2019} - -@article{Dongarra2009-na, - author = {Dongarra, Jack J}, - journal = {IBM Journal of Research and Development}, - pages = {3--4}, - title = {The evolution of high performance computing on system z}, - volume = 53, - year = 2009} - -@article{Ranganathan2011-dc, - author = {Ranganathan, Parthasarathy}, - journal = {Computer (Long Beach Calif.)}, - month = jan, - number = 1, - pages = {39--48}, - publisher = {Institute of Electrical and Electronics Engineers (IEEE)}, - title = {From microprocessors to nanostores: Rethinking data-centric systems}, - volume = 44, - year = 2011} - -@article{Ignatov2018-kh, - abstract = {Over the last years, the computational power of mobile devices - such as smartphones and tablets has grown dramatically, reaching - the level of desktop computers available not long ago. While - standard smartphone apps are no longer a problem for them, there - is still a group of tasks that can easily challenge even - high-end devices, namely running artificial intelligence - algorithms. In this paper, we present a study of the current - state of deep learning in the Android ecosystem and describe - available frameworks, programming models and the limitations of - running AI on smartphones. We give an overview of the hardware - acceleration resources available on four main mobile chipset - platforms: Qualcomm, HiSilicon, MediaTek and Samsung. - Additionally, we present the real-world performance results of - different mobile SoCs collected with AI Benchmark that are - covering all main existing hardware configurations.}, - author = {Ignatov, Andrey and Timofte, Radu and Chou, William and Wang, Ke and Wu, Max and Hartley, Tim and Van Gool, Luc}, - publisher = {arXiv}, - title = {{AI} Benchmark: Running deep neural networks on Android smartphones}, - year = 2018} - -@article{Sze2017-ak, - abstract = {Deep neural networks (DNNs) are currently widely used for - many artificial intelligence (AI) applications including - computer vision, speech recognition, and robotics. While - DNNs deliver state-of-the-art accuracy on many AI tasks, it - comes at the cost of high computational complexity. - Accordingly, techniques that enable efficient processing of - DNNs to improve energy efficiency and throughput without - sacrificing application accuracy or increasing hardware cost - are critical to the wide deployment of DNNs in AI systems. - This article aims to provide a comprehensive tutorial and - survey about the recent advances towards the goal of - enabling efficient processing of DNNs. Specifically, it will - provide an overview of DNNs, discuss various hardware - platforms and architectures that support DNNs, and highlight - key trends in reducing the computation cost of DNNs either - solely via hardware design changes or via joint hardware - design and DNN algorithm changes. It will also summarize - various development resources that enable researchers and - practitioners to quickly get started in this field, and - highlight important benchmarking metrics and design - considerations that should be used for evaluating the - rapidly growing number of DNN hardware designs, optionally - including algorithmic co-designs, being proposed in academia - and industry. The reader will take away the following - concepts from this article: understand the key design - considerations for DNNs; be able to evaluate different DNN - hardware implementations with benchmarks and comparison - metrics; understand the trade-offs between various hardware - architectures and platforms; be able to evaluate the utility - of various DNN design techniques for efficient processing; - and understand recent implementation trends and - opportunities.}, - archiveprefix = {arXiv}, - author = {Sze, Vivienne and Chen, Yu-Hsin and Yang, Tien-Ju and Emer, Joel}, - copyright = {http://arxiv.org/licenses/nonexclusive-distrib/1.0/}, - eprint = {1703.09039}, - month = mar, - primaryclass = {cs.CV}, - title = {Efficient processing of deep neural networks: A tutorial and survey}, - year = 2017} - -@inproceedings{lin2022ondevice, - author = {Lin, Ji and Zhu, Ligeng and Chen, Wei-Ming and Wang, Wei-Chen and Gan, Chuang and Han, Song}, - booktitle = {ArXiv}, - title = {On-Device Training Under 256KB Memory}, - year = {2022}} - -@article{lin2023awq, - author = {Lin, Ji and Tang, Jiaming and Tang, Haotian and Yang, Shang and Dang, Xingyu and Han, Song}, - journal = {arXiv}, - title = {AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration}, - year = {2023}} - -@inproceedings{wang2020apq, - author = {Wang, Tianzhe and Wang, Kuan and Cai, Han and Lin, Ji and Liu, Zhijian and Wang, Hanrui and Lin, Yujun and Han, Song}, - booktitle = {2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, - doi = {10.1109/CVPR42600.2020.00215}, - pages = {2075-2084}, - title = {APQ: Joint Search for Network Architecture, Pruning and Quantization Policy}, - year = {2020}, - Bdsk-Url-1 = {https://doi.org/10.1109/CVPR42600.2020.00215}} - -@inproceedings{Li2020Additive, - author = {Yuhang Li and Xin Dong and Wei Wang}, - booktitle = {International Conference on Learning Representations}, - title = {Additive Powers-of-Two Quantization: An Efficient Non-uniform Discretization for Neural Networks}, - url = {https://openreview.net/forum?id=BkgXT24tDS}, - year = {2020}, - Bdsk-Url-1 = {https://openreview.net/forum?id=BkgXT24tDS}} - -@article{janapa2023edge, - author = {Janapa Reddi, Vijay and Elium, Alexander and Hymel, Shawn and Tischler, David and Situnayake, Daniel and Ward, Carl and Moreau, Louis and Plunkett, Jenny and Kelcey, Matthew and Baaijens, Mathijs and others}, - journal = {Proceedings of Machine Learning and Systems}, - title = {Edge Impulse: An MLOps Platform for Tiny Machine Learning}, - volume = {5}, - year = {2023}} - -@article{zhuang2020comprehensive, - author = {Zhuang, Fuzhen and Qi, Zhiyuan and Duan, Keyu and Xi, Dongbo and Zhu, Yongchun and Zhu, Hengshu and Xiong, Hui and He, Qing}, - journal = {Proceedings of the IEEE}, - number = {1}, - pages = {43--76}, - publisher = {IEEE}, - title = {A comprehensive survey on transfer learning}, - volume = {109}, - year = {2020}} - -@article{zhuang_comprehensive_2021, - author = {Zhuang, Fuzhen and Qi, Zhiyuan and Duan, Keyu and Xi, Dongbo and Zhu, Yongchun and Zhu, Hengshu and Xiong, Hui and He, Qing}, - doi = {10.1109/JPROC.2020.3004555}, - file = {Zhuang et al. - 2021 - A Comprehensive Survey on Transfer Learning.pdf:/Users/alex/Zotero/storage/CHJB2WE4/Zhuang et al. - 2021 - A Comprehensive Survey on Transfer Learning.pdf:application/pdf}, - issn = {0018-9219, 1558-2256}, - journal = {Proceedings of the IEEE}, - language = {en}, - month = jan, - number = {1}, - pages = {43--76}, - title = {A {Comprehensive} {Survey} on {Transfer} {Learning}}, - url = {https://ieeexplore.ieee.org/document/9134370/}, - urldate = {2023-10-25}, - volume = {109}, - year = {2021}, - Bdsk-Url-1 = {https://ieeexplore.ieee.org/document/9134370/}, - Bdsk-Url-2 = {https://doi.org/10.1109/JPROC.2020.3004555}} - -@inproceedings{Norman2017TPUv1, - abstract = {Many architects believe that major improvements in cost-energy-performance must now come from domain-specific hardware. This paper evaluates a custom ASIC---called a Tensor Processing Unit (TPU) --- deployed in datacenters since 2015 that accelerates the inference phase of neural networks (NN). The heart of the TPU is a 65,536 8-bit MAC matrix multiply unit that offers a peak throughput of 92 TeraOps/second (TOPS) and a large (28 MiB) software-managed on-chip memory. The TPU's deterministic execution model is a better match to the 99th-percentile response-time requirement of our NN applications than are the time-varying optimizations of CPUs and GPUs that help average throughput more than guaranteed latency. The lack of such features helps explain why, despite having myriad MACs and a big memory, the TPU is relatively small and low power. We compare the TPU to a server-class Intel Haswell CPU and an Nvidia K80 GPU, which are contemporaries deployed in the same datacenters. Our workload, written in the high-level TensorFlow framework, uses production NN applications (MLPs, CNNs, and LSTMs) that represent 95\% of our datacenters' NN inference demand. Despite low utilization for some applications, the TPU is on average about 15X -- 30X faster than its contemporary GPU or CPU, with TOPS/Watt about 30X -- 80X higher. Moreover, using the CPU's GDDR5 memory in the TPU would triple achieved TOPS and raise TOPS/Watt to nearly 70X the GPU and 200X the CPU.}, - address = {New York, NY, USA}, - author = {Jouppi, Norman P. and Young, Cliff and Patil, Nishant and Patterson, David and Agrawal, Gaurav and Bajwa, Raminder and Bates, Sarah and Bhatia, Suresh and Boden, Nan and Borchers, Al and Boyle, Rick and Cantin, Pierre-luc and Chao, Clifford and Clark, Chris and Coriell, Jeremy and Daley, Mike and Dau, Matt and Dean, Jeffrey and Gelb, Ben and Ghaemmaghami, Tara Vazir and Gottipati, Rajendra and Gulland, William and Hagmann, Robert and Ho, C. Richard and Hogberg, Doug and Hu, John and Hundt, Robert and Hurt, Dan and Ibarz, Julian and Jaffey, Aaron and Jaworski, Alek and Kaplan, Alexander and Khaitan, Harshit and Killebrew, Daniel and Koch, Andy and Kumar, Naveen and Lacy, Steve and Laudon, James and Law, James and Le, Diemthu and Leary, Chris and Liu, Zhuyuan and Lucke, Kyle and Lundin, Alan and MacKean, Gordon and Maggiore, Adriana and Mahony, Maire and Miller, Kieran and Nagarajan, Rahul and Narayanaswami, Ravi and Ni, Ray and Nix, Kathy and Norrie, Thomas and Omernick, Mark and Penukonda, Narayana and Phelps, Andy and Ross, Jonathan and Ross, Matt and Salek, Amir and Samadiani, Emad and Severn, Chris and Sizikov, Gregory and Snelham, Matthew and Souter, Jed and Steinberg, Dan and Swing, Andy and Tan, Mercedes and Thorson, Gregory and Tian, Bo and Toma, Horia and Tuttle, Erick and Vasudevan, Vijay and Walter, Richard and Wang, Walter and Wilcox, Eric and Yoon, Doe Hyun}, - booktitle = {Proceedings of the 44th Annual International Symposium on Computer Architecture}, - doi = {10.1145/3079856.3080246}, - isbn = {9781450348928}, - keywords = {accelerator, neural network, MLP, TPU, CNN, deep learning, domain-specific architecture, GPU, TensorFlow, DNN, RNN, LSTM}, - location = {Toronto, ON, Canada}, - numpages = {12}, - pages = {1-12}, - publisher = {Association for Computing Machinery}, - series = {ISCA '17}, - title = {In-Datacenter Performance Analysis of a Tensor Processing Unit}, - url = {https://doi.org/10.1145/3079856.3080246}, - year = {2017}, - Bdsk-Url-1 = {https://doi.org/10.1145/3079856.3080246}} - -@article{Norrie2021TPUv2_3, - author = {Norrie, Thomas and Patil, Nishant and Yoon, Doe Hyun and Kurian, George and Li, Sheng and Laudon, James and Young, Cliff and Jouppi, Norman and Patterson, David}, - doi = {10.1109/MM.2021.3058217}, - journal = {IEEE Micro}, - number = {2}, - pages = {56-63}, - title = {The Design Process for Google's Training Chips: TPUv2 and TPUv3}, - volume = {41}, - year = {2021}, - Bdsk-Url-1 = {https://doi.org/10.1109/MM.2021.3058217}} - -@inproceedings{Jouppi2023TPUv4, - abstract = {In response to innovations in machine learning (ML) models, production workloads changed radically and rapidly. TPU v4 is the fifth Google domain specific architecture (DSA) and its third supercomputer for such ML models. Optical circuit switches (OCSes) dynamically reconfigure its interconnect topology to improve scale, availability, utilization, modularity, deployment, security, power, and performance; users can pick a twisted 3D torus topology if desired. Much cheaper, lower power, and faster than Infiniband, OCSes and underlying optical components are <5\% of system cost and <3\% of system power. Each TPU v4 includes SparseCores, dataflow processors that accelerate models that rely on embeddings by 5x--7x yet use only 5\% of die area and power. Deployed since 2020, TPU v4 outperforms TPU v3 by 2.1x and improves performance/Watt by 2.7x. The TPU v4 supercomputer is 4x larger at 4096 chips and thus nearly 10x faster overall, which along with OCS flexibility and availability allows a large language model to train at an average of ~60\% of peak FLOPS/second. For similar sized systems, it is ~4.3x--4.5x faster than the Graphcore IPU Bow and is 1.2x--1.7x faster and uses 1.3x--1.9x less power than the Nvidia A100. TPU v4s inside the energy-optimized warehouse scale computers of Google Cloud use ~2--6x less energy and produce ~20x less CO2e than contemporary DSAs in typical on-premise data centers.}, - address = {New York, NY, USA}, - articleno = {82}, - author = {Jouppi, Norm and Kurian, George and Li, Sheng and Ma, Peter and Nagarajan, Rahul and Nai, Lifeng and Patil, Nishant and Subramanian, Suvinay and Swing, Andy and Towles, Brian and Young, Clifford and Zhou, Xiang and Zhou, Zongwei and Patterson, David A}, - booktitle = {Proceedings of the 50th Annual International Symposium on Computer Architecture}, - doi = {10.1145/3579371.3589350}, - isbn = {9798400700958}, - keywords = {warehouse scale computer, embeddings, supercomputer, domain specific architecture, reconfigurable, TPU, large language model, power usage effectiveness, CO2 equivalent emissions, energy, optical interconnect, IPU, machine learning, GPU, carbon emissions}, - location = {Orlando, FL, USA}, - numpages = {14}, - publisher = {Association for Computing Machinery}, - series = {ISCA '23}, - title = {TPU v4: An Optically Reconfigurable Supercomputer for Machine Learning with Hardware Support for Embeddings}, - url = {https://doi.org/10.1145/3579371.3589350}, - year = {2023}, - Bdsk-Url-1 = {https://doi.org/10.1145/3579371.3589350}} - + title = {Deep {Class}-{Incremental} {Learning}: {A} {Survey}}, + shorttitle = {Deep {Class}-{Incremental} {Learning}}, + author = {Zhou, Da-Wei and Wang, Qi-Wei and Qi, Zhi-Hong and Ye, Han-Jia and Zhan, De-Chuan and Liu, Ziwei}, + year = 2023, + month = feb, + publisher = {arXiv}, + url = {http://arxiv.org/abs/2302.03648}, + urldate = {2023-10-26}, + note = {arXiv:2302.03648 [cs]}, + annote = {Comment: Code is available at https://github.com/zhoudw-zdw/CIL\_Survey/}, + language = {en}, + bdsk-url-1 = {http://arxiv.org/abs/2302.03648}, + file = {Zhou et al. - 2023 - Deep Class-Incremental Learning A Survey.pdf:/Users/alex/Zotero/storage/859VZG7W/Zhou et al. - 2023 - Deep Class-Incremental Learning A Survey.pdf:application/pdf}, + keywords = {Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning} +} +@inproceedings{zhou2018interpretable, + title = {Interpretable basis decomposition for visual explanation}, + author = {Zhou, Bolei and Sun, Yiyou and Bau, David and Torralba, Antonio}, + year = 2018, + booktitle = {Proceedings of the European Conference on Computer Vision (ECCV)}, + pages = {119--134} +} @misc{zhou2021analognets, + title = {AnalogNets: ML-HW Co-Design of Noise-robust TinyML Models and Always-On Analog Compute-in-Memory Accelerator}, + author = {Chuteng Zhou and Fernando Garcia Redondo and Julian B{\"u}chel and Irem Boybat and Xavier Timoneda Comas and S. R. Nandakumar and Shidhartha Das and Abu Sebastian and Manuel Le Gallo and Paul N. Whatmough}, + year = 2021, archiveprefix = {arXiv}, - author = {Chuteng Zhou and Fernando Garcia Redondo and Julian B{\"u}chel and Irem Boybat and Xavier Timoneda Comas and S. R. Nandakumar and Shidhartha Das and Abu Sebastian and Manuel Le Gallo and Paul N. Whatmough}, - eprint = {2111.06503}, - primaryclass = {cs.AR}, - title = {AnalogNets: ML-HW Co-Design of Noise-robust TinyML Models and Always-On Analog Compute-in-Memory Accelerator}, - year = 2021} - -@article{wearableinsulin, - article-number = {719}, - author = {Psoma, Sotiria D. and Kanthou, Chryso}, - doi = {10.3390/bios13070719}, - issn = {2079-6374}, - journal = {Biosensors}, - number = {7}, - pubmedid = {37504117}, - title = {Wearable Insulin Biosensors for Diabetes Management: Advances and Challenges}, - url = {https://www.mdpi.com/2079-6374/13/7/719}, - volume = {13}, - year = {2023}, - Bdsk-Url-1 = {https://www.mdpi.com/2079-6374/13/7/719}, - Bdsk-Url-2 = {https://doi.org/10.3390/bios13070719}} - -@article{glucosemonitor, - author = {Li, Jingzhen and Tobore, Igbe and Liu, Yuhang and Kandwal, Abhishek and Wang, Lei and Nie, Zedong}, - doi = {10.1109/JBHI.2021.3072628}, - journal = {IEEE Journal of Biomedical and Health Informatics}, - number = {9}, - pages = {3340-3350}, - title = {Non-invasive Monitoring of Three Glucose Ranges Based On ECG By Using DBSCAN-CNN}, - volume = {25}, - year = {2021}, - Bdsk-Url-1 = {https://doi.org/10.1109/JBHI.2021.3072628}} - -@article{plasma, - author = {Attia, Zachi and Sugrue, Alan and Asirvatham, Samuel and Ackerman, Michael and Kapa, Suraj and Friedman, Paul and Noseworthy, Peter}, - doi = {10.1371/journal.pone.0201059}, - journal = {PLOS ONE}, - month = {08}, - pages = {e0201059}, - title = {Noninvasive assessment of dofetilide plasma concentration using a deep learning (neural network) analysis of the surface electrocardiogram: A proof of concept study}, - volume = {13}, - year = {2018}, - Bdsk-Url-1 = {https://doi.org/10.1371/journal.pone.0201059}} - -@article{afib, - author = {Yutao Guo and Hao Wang and Hui Zhang and Tong Liu and Zhaoguang Liang and Yunlong Xia and Li Yan and Yunli Xing and Haili Shi and Shuyan Li and Yanxia Liu and Fan Liu and Mei Feng and Yundai Chen and Gregory Y.H. Lip and null null}, - doi = {10.1016/j.jacc.2019.08.019}, - journal = {Journal of the American College of Cardiology}, - number = {19}, - pages = {2365-2375}, - title = {Mobile Photoplethysmographic Technology to Detect Atrial Fibrillation}, - volume = {74}, - year = {2019}, - Bdsk-Url-1 = {https://doi.org/10.1016/j.jacc.2019.08.019}} - -@article{gaitathome, - author = {Yingcheng Liu and Guo Zhang and Christopher G. Tarolli and Rumen Hristov and Stella Jensen-Roberts and Emma M. Waddell and Taylor L. Myers and Meghan E. Pawlik and Julia M. Soto and Renee M. Wilson and Yuzhe Yang and Timothy Nordahl and Karlo J. Lizarraga and Jamie L. Adams and Ruth B. Schneider and Karl Kieburtz and Terry Ellis and E. Ray Dorsey and Dina Katabi}, - doi = {10.1126/scitranslmed.adc9669}, - eprint = {https://www.science.org/doi/pdf/10.1126/scitranslmed.adc9669}, - journal = {Science Translational Medicine}, - number = {663}, - pages = {eadc9669}, - title = {Monitoring gait at home with radio waves in Parkinson's disease: A marker of severity, progression, and medication response}, - url = {https://www.science.org/doi/abs/10.1126/scitranslmed.adc9669}, - volume = {14}, - year = {2022}, - Bdsk-Url-1 = {https://www.science.org/doi/abs/10.1126/scitranslmed.adc9669}, - Bdsk-Url-2 = {https://doi.org/10.1126/scitranslmed.adc9669}} - -@article{Chen2023, - author = {Chen, Emma and Prakash, Shvetank and Janapa Reddi, Vijay and Kim, David and Rajpurkar, Pranav}, - day = {06}, - doi = {10.1038/s41551-023-01115-0}, - issn = {2157-846X}, - journal = {Nature Biomedical Engineering}, - month = {Nov}, - title = {A framework for integrating artificial intelligence for clinical care with continuous therapeutic monitoring}, - url = {https://doi.org/10.1038/s41551-023-01115-0}, - year = {2023}, - Bdsk-Url-1 = {https://doi.org/10.1038/s41551-023-01115-0}} - -@article{Zhang2017, - author = {Zhang, Qingxue and Zhou, Dian and Zeng, Xuan}, - day = {06}, - doi = {10.1186/s12938-017-0317-z}, - issn = {1475-925X}, - journal = {BioMedical Engineering OnLine}, - month = {Feb}, - number = {1}, - pages = {23}, - title = {Highly wearable cuff-less blood pressure and heart rate monitoring with single-arm electrocardiogram and photoplethysmogram signals}, - url = {https://doi.org/10.1186/s12938-017-0317-z}, - volume = {16}, - year = {2017}, - Bdsk-Url-1 = {https://doi.org/10.1186/s12938-017-0317-z}} - -@misc{yik2023neurobench, - archiveprefix = {arXiv}, - author = {Jason Yik and Soikat Hasan Ahmed and Zergham Ahmed and Brian Anderson and Andreas G. Andreou and Chiara Bartolozzi and Arindam Basu and Douwe den Blanken and Petrut Bogdan and Sander Bohte and Younes Bouhadjar and Sonia Buckley and Gert Cauwenberghs and Federico Corradi and Guido de Croon and Andreea Danielescu and Anurag Daram and Mike Davies and Yigit Demirag and Jason Eshraghian and Jeremy Forest and Steve Furber and Michael Furlong and Aditya Gilra and Giacomo Indiveri and Siddharth Joshi and Vedant Karia and Lyes Khacef and James C. Knight and Laura Kriener and Rajkumar Kubendran and Dhireesha Kudithipudi and Gregor Lenz and Rajit Manohar and Christian Mayr and Konstantinos Michmizos and Dylan Muir and Emre Neftci and Thomas Nowotny and Fabrizio Ottati and Ayca Ozcelikkale and Noah Pacik-Nelson and Priyadarshini Panda and Sun Pao-Sheng and Melika Payvand and Christian Pehle and Mihai A. Petrovici and Christoph Posch and Alpha Renner and Yulia Sandamirskaya and Clemens JS Schaefer and Andr{\'e} van Schaik and Johannes Schemmel and Catherine Schuman and Jae-sun Seo and Sadique Sheik and Sumit Bam Shrestha and Manolis Sifalakis and Amos Sironi and Kenneth Stewart and Terrence C. Stewart and Philipp Stratmann and Guangzhi Tang and Jonathan Timcheck and Marian Verhelst and Craig M. Vineyard and Bernhard Vogginger and Amirreza Yousefzadeh and Biyan Zhou and Fatima Tuz Zohora and Charlotte Frenkel and Vijay Janapa Reddi}, - eprint = {2304.04640}, - primaryclass = {cs.AI}, - title = {NeuroBench: Advancing Neuromorphic Computing through Collaborative, Fair and Representative Benchmarking}, - year = {2023}} \ No newline at end of file + eprint = {2111.06503}, + primaryclass = {cs.AR} +} +@article{zhou2022photonic, + title = {Photonic matrix multiplication lights up photonic accelerator and beyond}, + author = {Zhou, Hailong and Dong, Jianji and Cheng, Junwei and Dong, Wenchan and Huang, Chaoran and Shen, Yichen and Zhang, Qiming and Gu, Min and Qian, Chao and Chen, Hongsheng and others}, + year = 2022, + journal = {Light: Science \& Applications}, + publisher = {Nature Publishing Group UK London}, + volume = 11, + number = 1, + pages = 30 +} +@inproceedings{zhu2018benchmarking, + title = {Benchmarking and analyzing deep neural network training}, + author = {Zhu, Hongyu and Akrout, Mohamed and Zheng, Bojian and Pelegris, Andrew and Jayarajan, Anand and Phanishayee, Amar and Schroeder, Bianca and Pekhimenko, Gennady}, + year = 2018, + booktitle = {2018 IEEE International Symposium on Workload Characterization (IISWC)}, + pages = {88--100}, + organization = {IEEE} +} +@article{zhuang_comprehensive_2021, + title = {A {Comprehensive} {Survey} on {Transfer} {Learning}}, + author = {Zhuang, Fuzhen and Qi, Zhiyuan and Duan, Keyu and Xi, Dongbo and Zhu, Yongchun and Zhu, Hengshu and Xiong, Hui and He, Qing}, + year = 2021, + month = jan, + journal = {Proceedings of the IEEE}, + publisher = {IEEE}, + volume = 109, + number = 1, + pages = {43--76}, + doi = {10.1109/JPROC.2020.3004555}, + issn = {0018-9219, 1558-2256}, + url = {https://ieeexplore.ieee.org/document/9134370/}, + urldate = {2023-10-25}, + language = {en}, + file = {Zhuang et al. - 2021 - A Comprehensive Survey on Transfer Learning.pdf:/Users/alex/Zotero/storage/CHJB2WE4/Zhuang et al. - 2021 - A Comprehensive Survey on Transfer Learning.pdf:application/pdf}, + bdsk-url-1 = {https://ieeexplore.ieee.org/document/9134370/}, + bdsk-url-2 = {https://doi.org/10.1109/JPROC.2020.3004555}, + file = {Zhuang et al. - 2021 - A Comprehensive Survey on Transfer Learning.pdf:/Users/alex/Zotero/storage/CHJB2WE4/Zhuang et al. - 2021 - A Comprehensive Survey on Transfer Learning.pdf:application/pdf} +} +@article{zhuang2020comprehensive, + title = {A comprehensive survey on transfer learning}, + author = {Zhuang, Fuzhen and Qi, Zhiyuan and Duan, Keyu and Xi, Dongbo and Zhu, Yongchun and Zhu, Hengshu and Xiong, Hui and He, Qing}, + year = 2020, + journal = {Proceedings of the IEEE}, + publisher = {IEEE}, + volume = 109, + number = 1, + pages = {43--76} +} \ No newline at end of file diff --git a/sustainable_ai.qmd b/sustainable_ai.qmd index 8573a4e8..881980d9 100644 --- a/sustainable_ai.qmd +++ b/sustainable_ai.qmd @@ -1,84 +1,617 @@ -# Sustainable AI +# Sustainable AI {#sustainable-ai} ![_DALL·E 3 Prompt: 3D illustration on a light background of a sustainable AI network interconnected with a myriad of eco-friendly energy sources. The AI actively manages and optimizes its energy from sources like solar arrays, wind turbines, and hydro dams, emphasizing power efficiency and performance. Deep neural networks spread throughout, receiving energy from these sustainable resources._](./images/cover_sustainable_ai.png) -## Introduction +::: {.callout-tip} -Explanation: In this introductory section, we elucidate the significance of sustainability in the context of AI, emphasizing the necessity to address environmental, economic, and social dimensions to build resilient and sustainable AI systems. +## Learning Objectives -- Importance of sustainability in AI -- Sustainability dimensions: environmental, economic, and social -- Overview of challenges and opportunities +* Understand the various aspects of AI's environmental impact, including energy consumption, carbon emissions, electronic waste, and biodiversity effects. +* Learn about methods and best practices for developing sustainable AI systems +* Appreciate the importance of taking a lifecycle perspective when evaluating and addressing the sustainability of AI systems. +* Recognize the roles various stakeholders like researchers, corporations, policymakers and end users play in furthering responsible and sustainable AI progress. +* Learn about specific frameworks, metrics and tools aimed at enabling greener AI development. +* Appreciate real-world case studies like Google's 4M efficiency practices that showcase how organizations are taking tangible steps to improve AI's environmental record -## Energy Efficiency of AI Models +::: -Explanation: This section addresses the pressing issue of high energy consumption associated with AI models, offering insights into techniques for creating energy-efficient AI models which are not only economical but also environmentally friendly. +## Introduction {#introduction} -- Energy consumption patterns of AI models -- Techniques for improving energy efficiency -- Case studies of energy-efficient AI deployments +The rapid advancements in artificial intelligence (AI) and machine learning (ML) have led to many beneficial applications and optimizations for performance efficiency. However, the remarkable growth of AI comes with a significant, yet often overlooked cost: its environmental impact. The most recent report released by the IPCC, the international body leading scientific assessments of climate change and its impacts, emphasized the pressing importance of tackling climate change. Without immediate efforts to decrease global $\textrm{CO}_2$ emissions by at least 43 percent before 2030, we exceed global warming of 1.5 degrees celsius [@lecocq2022mitigation]. This could initiate positive feedback loops pushing temperatures even higher. Next to environmental issues, the United Nations recognized [17 Sustainable Development Goals (SDGs)](https://sdgs.un.org/goals), in which AI can play an important role, and vice versa, play an important role in the development of AI systems. As the field continues expanding, considering sustainability is crucial. -## Responsible Resource Utilization +AI systems, particularly large language models like [GPT-3](https://openai.com/blog/gpt-3-apps/) and computer vision models like [DALL-E 2](https://openai.com/dall-e-2/), require massive amounts of computational resources for training. For example, GPT-3 was estimated to consume 1,300 megawatt-hours of electricity, which is equal to 1,450 average U.S. households in an entire month [@maslej2023artificial], or put another way it consumed enough energy to supply an average U.S. household for 120 years! This immense energy demand stems primarily from power-hungry data centers with servers running intense computations to train these complex neural networks for days or weeks. -Explanation: Here, we delve into strategies for responsible resource utilization in AI, discussing how optimizing resource allocation can lead to more sustainable and cost-effective AI systems. +Current estimates indicate that the carbon emissions produced from developing a single sophisticated AI model can equal the emissions over the lifetime of five standard gasoline-powered vehicles [@strubellEnergyPolicyConsiderations2019]. A significant portion of the electricity presently consumed by data centers is generated from nonrenewable sources such as coal and natural gas, resulting in data centers contributing around [1% of total worldwide carbon emissions](https://www.iea.org/energy-system/buildings/data-centres-and-data-transmission-networks). This is comparable to the emissions from the entire airline sector. This immense carbon footprint demonstrates the pressing need to transition to renewable power sources such as solar and wind to operate AI development. -- Resource allocation and management in AI -- Reducing resource wastage -- Resource optimization techniques and tools -- Explain resource difference between big / small systems +Additionally, even small-scale AI systems deployed to edge devices as part of TinyML have environmental impacts that should not be ignored [@prakashTinyMLSustainableAssessing2023]. The specialized hardware required for AI has an environmental toll from natural resource extraction and manufacturing. GPUs, CPUs, and chips like TPUs depend on rare earth metals whose mining and processing generate substantial pollution. The production of these components also has its energy demands. Furthermore, the process of collecting, storing, and preprocessing data used to train both small- and large-scale models comes with environmental costs, which further exacerbates the sustainability implications of ML systems. -## E-Waste Management +Thus, while AI promises innovative breakthroughs in many fields, sustaining progress requires addressing its sustainability challenges. AI can continue advancing responsibly by optimizing the efficiency of models, exploring alternative specialized hardware and renewable energy sources for data centers, and tracking the overall environmental impact. -Explanation: This segment explores the problem of electronic waste generated by AI components, suggesting guidelines and best practices for reducing e-waste and promoting recycling and reusing initiatives. +## Social and Ethical Responsibility {#social-and-ethical-responsibility} -- Overview of e-waste generated by AI components -- Best practices for e-waste management -- Promoting recycling and reuse in AI systems -- Discuss tinyML e-waste from CACM +The environmental impact of AI is not just a technical issue but an ethical and social one as well. As AI becomes more integrated into our lives and industries, its sustainability becomes increasingly critical. -## Carbon Footprint Reduction +### Ethical Considerations {#ethical-considerations} -Explanation: In this section, readers will learn about the carbon footprint associated with AI operations and the methods to mitigate it, contributing to a greener and more sustainable AI ecosystem. +The scale of AI's environmental footprint raises profound ethical questions about the responsibilities of AI developers and companies to minimize their carbon emissions and energy usage. As the creators of AI systems and technologies that can have sweeping global impacts, developers have an ethical obligation to consciously integrate environmental stewardship into their design process, even if sustainability comes at the cost of some efficiency gains. -- Assessing the carbon footprint of AI operations -- Strategies for carbon footprint reduction -- Discuss how edge/tinyML might help address issues -- Carbon offset initiatives in AI +There is a clear and present need for us to have open and honest conversations about AI's environmental tradeoffs earlier in the development lifecycle. Researchers should feel empowered to voice concerns if organizational priorities do not align with ethical goals, as in the case of the [open letter to pause giant AI experiments](https://futureoflife.org/open-letter/pause-giant-ai-experiments/). -## Sustainable Embedded ML +Additionally, there is increasing need for AI companies to scrutinize their contributions to climate change and environmental harm. Large tech firms are responsible for the cloud infrastructure, data center energy demands, and resource extraction required to power today's AI. Leadership should assess if organizational values and policies promote sustainability, from hardware manufacturing through model training pipelines. -Explanation: The focus here is on the full footprint, embodied and carbon footprint, which are the backbone of sustainability, providing insights into how the devices can be designed or modified to be more sustainable +Furthermore, voluntary self-regulation may not be enough – governments may need to introduce new regulations aimed at sustainable AI standards and practices if we hope to curb the projected energy explosion of ever-larger models. Reported metrics like compute usage, carbon footprint, and efficiency benchmarks could help hold organizations accountable. -- Read through the tinyML sustainability paper +Through ethical principles, company policies, and public rules, AI technologists and corporations have a profound duty to our planet to ensure the responsible and sustainable advancement of technology positioned to transform modern society radically. We owe it to future generations to get this right. -## Community Engagement and Collaboration +### Long-term Sustainability {#long-term-sustainability} -Explanation: This section accentuates the role of community engagement and collaboration in fostering AI sustainability, presenting ways in which a collaborative approach can help in sharing knowledge and resources for sustainable AI development. +The massive projected expansion of AI raises urgent concerns about its long-term sustainability. As AI software and applications rapidly increase in complexity and usage across industries, demand for computing power and infrastructure will skyrocket exponentially in the coming years. -- Community-driven sustainability initiatives -- Collaborative research and development -- Public-private partnerships for sustainable AI +To put the scale of projected growth in perspective, the total computing capacity required for training AI models saw an astonishing 350,000x increase from 2012 to 2019 [@schwartz2020green]. Researchers forecast over an order of magnitude growth each year moving forward as personalized AI assistants, autonomous technology, precision medicine tools, and more are developed. Similar trends are estimated for embedded ML systems, with an estimated 2.5 billion AI-enabled edge devices being deployed by 2030. -## Policy Frameworks and Regulations +Managing this expansion level requires software and hardware-focused breakthroughs in efficiency and renewable integration from AI engineers and scientists. On the software side, novel techniques in model optimization, distillation, pruning, low-precision numerics, knowledge sharing between systems, and other areas must become widespread best practices to curb energy needs. For example, realizing even a 50% reduced computational demand per capability doubling would have massive compounding on total energy. -Explanation: This segment emphasizes the necessity for robust policy frameworks and regulations to govern AI sustainability, highlighting global efforts and initiatives that are steering the path towards a sustainable AI future. +On the hardware infrastructure side, due to increasing costs of data transfer, storage, cooling, and space, continuing today's centralized server farm model at data centers is likely infeasible long-term [@lannelongueGreenAlgorithmsQuantifying2020]. Exploring alternative decentralized computing options around "edge AI" on local devices or within telco networks can alleviate scaling pressures on power-hungry hyperscale data centers. Likewise, the shift towards carbon-neutral, hybrid renewable energy sources powering leading cloud provider data centers worldwide will be essential. -- Existing policy frameworks for AI sustainability -- International initiatives and collaborations -- Future directions in policy and regulation +### AI for Environmental Good {#ai-for-environmental-good} -## Future Trends in AI Sustainability +While much focus goes on AI's sustainability challenges, these powerful technologies provide unique solutions to combat climate change and drive environmental progress. For example, ML can continuously optimize smart power grids to improve renewable integration and electricity distribution efficiency across networks [@zhang2018review]. Models can ingest the real-time status of a power grid and weather forecasts to allocate and shift sources responding to supply and demand. -Explanation: Here, we discuss anticipated trends in AI sustainability, projecting how evolving technologies and methodologies might shape the sustainability landscape of AI in the coming years. +Fine-tuned neural networks have also proven remarkably effective at next-generation [weather forecasting](https://deepmind.google/discover/blog/graphcast-ai-model-for-faster-and-more-accurate-global-weather-forecasting/) [@lam2023learning] and climate modeling [@kurth2023fourcastnet]. They can rapidly analyze massive volumes of climate data to boost extreme event preparation and resource planning for hurricanes, floods, droughts and more. Climate researchers have achieved state-of-the-art storm path accuracy by combining AI simulations with traditional numerical models. -- Anticipated technological advancements -- Role of AI in promoting global sustainability -- Challenges and opportunities ahead +AI also enables better tracking of biodiversity [@silvestro2022improving], wildlife [@schwartzDeploymentEmbeddedEdgeAI2021], [ecosystems](https://blogs.nvidia.com/blog/conservation-ai-detects-threats-to-endangered-species/#:~:text=The%20Conservation%20AI%20platform%20%E2%80%94%20built,of%20potential%20threats%20via%20email), and illegal deforestation using drones and satellite feeds. Computer vision algorithms can automate species population estimates and habitat health assessments over huge untracked regions. These capabilities provide conservationists with powerful tools for combating poaching [@bondi2018spot], reducing species extinction risks, and understanding ecological shifts. -## Conclusion +Targeted investment into AI applications for environmental sustainability, cross-sector data sharing, and model accessibility can profoundly accelerate solutions to pressing ecological issues. Emphasizing AI for social good steers innovation in cleaner directions, guiding these world-shaping technologies towards ethical and responsible development. -Explanation: The closing section encapsulates the key discussions and insights presented throughout the chapter, fostering a deep-seated understanding of the necessity and approaches for AI sustainability. +### Case Study -- Recap of key insights and discussions -- The road ahead: fostering sustainability in AI -- Encouraging innovation and research in AI sustainability +Google's data centers are foundational to powering products like Search, Gmail, and YouTube used by billions daily. However, keeping the vast server farms up and running requires substantial energy, particularly for vital cooling systems. Google continuously strives to enhance efficiency across operations. Yet progress was proving difficult through traditional methods alone considering the complex, custom dynamics involved. This challenge prompted an ML breakthrough yielding potential savings. + +After over a decade of optimizing data center design, inventing energy-efficient computing hardware, and securing renewable energy sources, [Google brought DeepMind scientists to unlock further advances](https://blog.google/outreach-initiatives/environment/deepmind-ai-reduces-energy-used-for/). The AI experts faced intricate factors surrounding the functioning of industrial cooling apparatuses. Equipment like pumps and chillers interact nonlinearly, while external weather and internal architectural variables also change. Capturing this complexity confounded rigid engineering formulas and human intuition. + +The DeepMind team leveraged Google's extensive historical sensor data detailing temperatures, power draw, and other attributes as training inputs. They built a flexible system based on neural networks to model the relationships and predict optimal configurations, minimizing power usage effectiveness (PUE) [@barroso2019datacenter]; PUE is the standard measurement for gauging how efficiently a data center uses energy—it gives the proportion of total facility power consumed divided by the power directly used for computing operations. When tested live, the AI system delivered remarkable gains beyond prior innovations, lowering cooling energy by 40% for a 15% drop in total PUE, a new site record. The generalizable framework learned cooling dynamics rapidly across shifting conditions that static rules could not match. The breakthrough highlights AI's rising role in transforming modern tech and enabling a sustainable future. + +## Energy Consumption {#energy-consumption} + +### Understanding Energy Needs {#understanding-energy-needs} + +In the rapidly evolving field of AI, understanding the energy needs for training and operating AI models is crucial. With AI entering widespread use in many new fields [@ai_health_rise, @data_centers_wheels], the demand for AI enabled devices and data centers is expected to explode. This understanding helps us grasp why AI, particularly deep learning, is often labeled as energy-intensive. + +#### Energy Requirements for AI Training {#energy-requirements-for-ai-training} + +![The performance of the language model improves smoothly with model size, dataset size and the amount of compute used for training [@scaling_laws_NLM].](images/sustainable_ai/model_scaling.png){#fig-scaling-laws} + +The training of complex AI systems like large deep learning models can demand startlingly high levels of computing power--with profound energy implications. Consider OpenAI’s state-of-the-art language model GPT-3 as a prime example. This system pushes the frontiers of text generation through algorithms trained on massive datasets, yet the energy GPT-3 consumed for a single training cycle could rival an [entire small town’s monthly usage](https://www.washington.edu/news/2023/07/27/how-much-energy-does-chatgpt-use/). In recent years, these generative AI models have gained increasing popularity, leading to an increased number of models being trained. Next to the increased number of models, the number of parameters in these models is likely to increase as well. Research shows that increasing the model size, dataset size and compute used for training improves performance smoothly with no signs of saturation [@scaling_laws_NLM], as seen in @fig-scaling-laws. + +What drives such immense requirements? During training, models like GPT-3 essentially learn their capabilities by continuously processing huge volumes of data to adjust internal parameters. The processing capacity that enables AI’s rapid advances also contributes to surging energy usage, especially as datasets and models balloon in size. In fact, GPT-3 highlights a steady trajectory in the field where each leap in AI’s sophistication traces back to ever more substantial computational power and resources. Its predecessor GPT-2 required 10x less training compute being only 1.5 billion parameters; a difference now dwarfed by magnitudes as GPT-3 comprises 175 billion parameters. Sustaining this trajectory toward increasingly capable AI therefore raises energy and infrastructure provision challenges ahead. + +#### Operational Energy Use {#operational-energy-use} + +The development and training of AI models requires immense amounts of data, computing power, and energy. However, the deployment and operation of those models also incurs significant recurrent resource costs over time. AI systems are now integrated across various industries and applications, and entering daily lives of an increasing demographic. Their cumulative operational energy and infrastructure impacts could eclipse that of the upfront model training. + +This concept is reflected in the demand of training and inference hardware, in datacenters and on the edge. Inference refers to the actual usage of a trained model to make predictions or decisions on real-world data. According to a [recent McKinsey analysis](https://www.mckinsey.com/~/media/McKinsey/Industries/Semiconductors/Our%20Insights/Artificial%20intelligence%20hardware%20New%20opportunities%20for%20semiconductor%20companies/Artificial-intelligence-hardware.ashx), the need for advanced systems to train ever-larger models is rapidly growing. However, inference computations already make up a dominant and increasing portion of total AI workloads, as shown in @fig-mckinsey. Running real-time inference with trained models--whether for image classification, speech recognition, or predictive analytics--invariably demands computing hardware like servers and chips. But even a model handling thousands of facial recognition requests or natural language queries daily is dwarfed by massive platforms like Meta. Where inference on millions of photos and videos shared on social media, the infrastructure energy requirements continue to scale! + +![At both the data centers and the edge, demand for training and inference hardware is growing.](images/sustainable_ai/mckinsey_analysis.png ""){#fig-mckinsey} + +Algorithms powering AI-enabled smart assistants, automated warehouses, self-driving vehicles, tailored healthcare, and more have marginal individual energy footprints. However, the projected proliferation of these technologies could add hundreds of millions of endpoints running AI algorithms continually, causing the scale of their collective energy requirements to surge. Current efficiency gains struggle to counterbalance this sheer growth. + +AI is expected to see an [annual growth rate of 37.3% between 2023 and 2030](https://www.forbes.com/advisor/business/ai-statistics/). Yet applying the same growth rate to operational compute could multiply annual AI energy needs up to 1000 times by 2030. So while model optimization tackles one facet, responsible innovation must also consider total lifecycle costs at global deployment scales that were unfathomable just years ago but now pose infrastructure and sustainability challenges ahead. + +### Data Centers and Their Impact {#data-centers-and-their-impact} + +The impact of data centers on the energy consumption of AI systems is a topic of increasing importance, as the demand for AI services grows. These facilities, while crucial for the advancement and deployment of AI, contribute significantly to its energy footprint. + +#### Scale {#scale} + +Data centers are the essential workhorses enabling the recent computational demands of advanced AI systems. For example, leading providers like Meta operate massive data centers spanning up to the [size of multiple football fields](https://tech.facebook.com/engineering/2021/8/eagle-mountain-data-center/), housing hundreds of thousands of high-capacity servers optimized for parallel processing and data throughput. + +These massive facilities provide the infrastructure for training complex neural networks on vast datasets--for instance, based on [leaked information](https://www.semianalysis.com/p/gpt-4-architecture-infrastructure), OpenAI's language model GPT-4 was trained on Azure data centers packing over 25,000 Nvidia A100 GPUs, used continuously for over 90 to 100 days. + +Additionally, real-time inference for consumer AI applications at scale is only made possible by leveraging the server farms inside data centers. Services like Alexa, Siri and Google Assistant process billions of voice requests per month from users globally by relying on data center computing for low-latency response. Going forward, expanding cutting-edge use cases like self-driving vehicles, precision medicine diagnostics, and accurate climate forecasting models require significant computational resources, obtained by tapping into vast on-demand cloud computing resources from data centers. For some emerging applications like autonomous cars, there are harsh latency and bandwidth constraints. Locating data center-level compute power on the edge rather than the cloud will be necessary. + +MIT research prototypes have shown trucks and cars with on-board hardware performing real-time AI processing of sensor data equivalent to small data centers [@data_centers_wheels]. These innovative “data centers on wheels” demonstrate how vehicles like self-driving trucks may need embedded data center-scale compute on board to achieve millisecond system latency for navigation, though still likely supplemented by wireless 5G connectivity to more powerful cloud data centers. + +The bandwidth, storage, and processing capacities required for enabling this future technology at scale will depend heavily on continuing data center infrastructure advancement alongside AI algorithmic innovations. + +#### Energy Demand {#energy-demand} + +The energy demand of data centers can roughly be divided into 4 components. Infrastructure, network, storage and servers. In @fig-energydemand, we see that the data center infrastructure, which includes aspects such as cooling, lighting and controls, and the servers, responsible for the compute, use the majority of the total energy budget.[@USA_energy] In this section, we break down the energy demand for the servers and the infrastructure. For the latter, the focus is laid on the cooling systems, as cooling is the dominant factor in energy consumption in the infrastructure. + +![Infrastructure and the servers consume the most energy in a datacenter.](images/sustainable_ai/energy_datacenter.png){#fig-energydemand} + +##### Servers {#servers} + +The increase in energy consumption of data centers stems mainly from exponentially growing AI computing requirements. NVIDIA DGX H100 machines that are optimized for deep learning can draw up to [10.2 kW at peak](https://docs.nvidia.com/dgx/dgxh100-user-guide/introduction-to-dgxh100.html). Leading providers operate data centers with hundreds to thousands of these power-hungry DGX nodes networked to train the latest AI models. For example, the supercomputer developed for OpenAI is a single system with more than 285,000 CPU cores, 10,000 GPUs and 400 gigabits per second of network connectivity for each GPU server. + +The intensive computations needed across an entire facility’s densely packed fleet and supporting hardware result in data centers drawing tens of megawatts around the clock. Overall, advancing AI algorithms continue to expand data center energy consumption as more DGX nodes get deployed to keep pace with projected growth in demand for AI compute resources over the coming years. + +##### Cooling Systems {#cooling-systems} + +To keep the beefy servers fed at peak capacity and cool, data centers require tremendous cooling capacity to counteract the heat produced by densely packed servers, networking equipment, and other hardware running computationally-intensive workloads without pause. With large data centers packing thousands of server racks operating at full tilt, massive industrial-scale cooling towers and chillers are required, using energy amounting to 30-40% of the total data center electricity footprint [@dayarathna2015data]. Consequently, companies are looking for alternative methods of cooling. For example, Microsoft’s data center in Ireland leverages a nearby fjord to exchange heat [using over half a million gallons of seawater daily](https://local.microsoft.com/communities/emea/dublin/). + +Recognizing the importance of energy-efficient cooling, there have been innovations aimed at reducing this energy demand. Techniques like free cooling, which uses outside air or water sources when conditions are favorable, and the use of AI to optimize cooling systems, are examples of how the industry is adapting. These innovations not only reduce energy consumption but also lower operational costs and lessen the environmental footprint. However, exponential increases in AI model complexity continue to demand more servers and acceleration hardware operating at higher utilization, translating to rising heat generation and ever greater energy used solely for cooling purposes. + +#### The Environmental Impact {#the-environmental-impact} + +The environmental impact of data centers is not only caused by direct energy consumption of the datacenter itself [@USA_footprint]. The operation of data centers involves the supply of treated water to the datacenter and the discharge of wastewater from the datacenter. Water and wastewater facilities are major electricity consumers. + +Next to electricity usage, there are many more aspects to the environmental impacts of these data centers. The water usage of the data centers can lead to water scarcity issues, increased water treatment needs and proper wastewater discharge infrastructure. Also raw materials required for construction and network transmission pose considerable impacts on the environment. Finally, components in data centers need to be upgraded and maintained. Where almost 50 percent of servers were refreshed within 3 years of usage, refresh cycles have shown to slow down [@uptime]. Still, this generates a significant amount of e-waste which can be hard to recycle. + +## Carbon Footprint {#carbon-footprint} + +The massive electricity demands of data centers can lead to significant environmental externalities absent an adequate renewable power supply. Many facilities rely heavily on non-renewable energy sources like coal and natural gas. For example, data centers are estimated to produce up to [2% of total global $\textrm{CO}_2$ emissions](https://www.independent.co.uk/climate-change/news/global-warming-data-centres-to-consume-three-times-as-much-energy-in-next-decade-experts-warn-a6830086.html) which is [closing the gap with the airline industry](@https://www.computerworld.com/article/3431148/why-data-centres-are-the-new-frontier-in-the-fight-against-climate-change.html). As mentioned in previous sections, the computational demands of AI are set to increase. The emissions of this surge are threefold. First, data centers are projected to increase in size [@EnergyCons_Emission]. Secondly, emissions during training are set to increase significantly [@Carbon_LNN]. Thirdly, inference calls to these models are set to increase dramatically as well. + +Without action, this exponential demand growth risks ratcheting up the carbon footprint of data centers further to unsustainable levels. Major providers have pledged carbon neutrality and committed funds to secure clean energy, but progress remains incremental compared to overall industry expansion plans. More radical grid decarbonization policies and renewable energy investments may prove essential to counteracting the climate impact of the coming tide of new data centers aimed at supporting the next generation of AI. + +### Definition and Significance {#definition-and-significance} + +The concept of a 'carbon footprint' has emerged as a key metric. This term refers to the total amount of greenhouse gasses, particularly carbon dioxide, that are emitted directly or indirectly by an individual, organization, event, or product. These emissions significantly contribute to the greenhouse effect, which in turn accelerates global warming and climate change. The carbon footprint is measured in terms of carbon dioxide equivalents ($\textrm{CO}_2$e), allowing for a comprehensive account that includes various greenhouse gasses and their relative impact on the environment. Examples of this as applied to large-scale ML tasks is shown in @fig-model_carbonfootprint. + +![The carbon footprint of large-scale ML tasks. It includes various models running at Meta. The emissions are categorized by offline training, online training, and inference [@wu2022sustainable].](images/sustainable_ai/model_carbonfootprint.png){#fig-model_carbonfootprint} + +The consideration of the carbon footprint is especially important in the field of AI. AI's rapid advancement and integration into various sectors have brought its environmental impact into sharp focus. AI systems, particularly those involving intensive computations like deep learning and large-scale data processing, are known for their substantial energy demands. This energy, often drawn from power grids, may still predominantly rely on fossil fuels, leading to significant greenhouse gas emissions. + +Take, for example, the training of large AI models such as GPT-3 or complex neural networks. These processes require immense computational power, typically provided by data centers. The energy consumption associated with operating these centers, particularly for such high-intensity tasks, results in notable greenhouse gas emissions. Studies have highlighted that training a single AI model can generate carbon emissions comparable to that of the lifetime emissions of multiple cars, shedding light on the environmental cost of developing advanced AI technologies [@dayarathna2015data]. @fig-carboncars shows a comparison from lowest to highest carbon footprints, starting with a roundtrip flight between NY and SF, human life average per year, American life average per year, US car including fuel over a lifetime, and a Transformer model with neural architecture search, which has the highest footprint. + +![Carbon footprint of NLP model in lbs of $\textrm{CO}_2$ equivalent. Source: [@dayarathna2015data]](images/sustainable_ai/carbon_benchmarks.png){#fig-carboncars} + +Moreover, the carbon footprint of AI extends beyond the operational phase. The entire lifecycle of AI systems, including the manufacturing of computing hardware, the energy used in data centers for cooling and maintenance, and the disposal of electronic waste, contributes to their overall carbon footprint. Some of which we have discussed earlier and we will discuss the waste aspects later on in this chapter. + +### The Need for Awareness and Action {#the-need-for-awareness-and-action} + +Understanding the carbon footprint of AI systems is crucial for several reasons. Primarily, it is a step towards mitigating the impacts of climate change. As AI continues to grow and permeate different aspects of our lives, its contribution to global carbon emissions becomes a significant concern. Awareness of these emissions can inform decisions made by developers, businesses, policymakers, and even ML engineers and scientists like us to ensure a balance between technological innovation and environmental responsibility. + +Furthermore, this understanding stimulates the drive towards 'Green AI' [@schwartz2020green]. This approach focuses on developing AI technologies that are efficient, powerful, and environmentally sustainable. It encourages the exploration of energy-efficient algorithms, the use of renewable energy sources in data centers, and the adoption of practices that reduce the overall environmental impact of AI. + +In essence, the carbon footprint is an essential consideration in developing and applying AI technologies. As AI evolves and its applications become more widespread, managing its carbon footprint is key to ensuring that this technological progress aligns with the broader environmental sustainability goals. + +### Estimating the AI Carbon Footprint {#estimating-the-ai-carbon-footprint} + +In understanding AI's environmental impact, estimating AI systems' carbon footprint is a critical step. This involves analyzing the various elements contributing to emissions throughout the lifecycle of AI technologies and employing specific methodologies to quantify these emissions accurately. Many different methods for quantifying these carbon emissions of ML have been proposed. + +The carbon footprint of AI encompasses several key elements, each contributing to the overall environmental impact. First, energy is consumed during AI model training and operational phases. The source of this energy heavily influences the carbon emissions. Once trained, these models, depending on their application and scale, continue to consume electricity during operation. Next to energy considerations, the hardware used stresses the environment as well. + +The carbon footprint varies significantly based on the energy sources used. The composition of the sources providing the energy used in the grid varies widely with geographical regions, and even with time in a single day! For example, in the USA, [roughly 60 percent of the total energy supply is still covered by fossil fuels](https://www.eia.gov/tools/faqs/faq.php?id=427&t=3). The remaining 40 percent is roughly equally covered by nuclear and renewable energy sources. These fractions are not constant throughout the day. As the production of renewable energy usually relies on environmental factors, such as solar radiation and pressure fields, they do not provide a constant source of energy. + +The variability of renewable energy production has been an ongoing challenge in the widespread use of these sources. Looking at @fig-energyprod, which shows data for the European grid, we see that it is not yet possible to produce the required amount of energy throughout the entire day. While solar energy peaks in the middle of the day, wind energy shows two distinct peaks in the mornings and evenings. Currently, to supply the lack of energy during times where renewable energy does not meet requirements, we rely on fossil and coal based energy generation methods. + +To enable constant use of renewable energy sources, innovation in energy storage solutions is required. Base energy load is currently met with nuclear energy. This constant energy source does not directly emit carbon emissions, but is too slow to accommodate for the variability of renewable energy sources. Tech companies such as Microsoft have shown interest in nuclear energy sources [to power their data centers](https://www.bloomberg.com/news/newsletters/2023-09-29/microsoft-msft-sees-artificial-intelligence-and-nuclear-energy-as-dynamic-duo). As the demand of data centers is more constant than the demand of regular households, nuclear energy could be used as a dominant source of energy. + +![Energy is supplied by various sources which vary throughout the day. Renewable energy production shows high variability with time. Source: [Energy Charts](https://www.energy-charts.info/?l=en&c=DE).](images/sustainable_ai/europe_energy_grid.png){#fig-energyprod} + +Additionally, the manufacturing and disposal of AI hardware add to the carbon footprint. The production of specialized computing devices, such as GPUs and CPUs, is an energy- and resource-intensive process. This phase often relies on energy sources that contribute to greenhouse gas emissions. The manufacturing process of the electronics industry has been identified as one of the big eight supply chains, responsible for more than 50 percent of total global emissions [@weforum]. Furthermore, the end-of-life disposal of this hardware, which can lead to electronic waste, also has environmental implications. As mentioned before, servers currently have a refresh cycle of roughly 3 to 5 years. Of this e-waste, currently [only 17.4 percent is properly collected and recycled.](https://www.genevaenvironmentnetwork.org/resources/updates/the-growing-environmental-risks-of-e-waste/) The carbon emissions of this e-waste has shown an increase of more than 50 percent between 2014 and 2020 [@e_waste]. + +As is clear from the above, a proper Life Cycle Analysis is necessary to portray all relevant aspects of the emissions caused by AI. Another method is carbon accounting, which quantifies the amount of carbon dioxide emissions directly and indirectly associated with AI operations. This measurement is typically in terms of $\textrm{CO}_2$ equivalents, allowing for a standardized way of reporting and assessing emissions. + +## Beyond Carbon Footprint {#beyond-carbon-footprint} + +The current focus on reducing the carbon emissions and energy consumption of AI systems addresses one crucial aspect of sustainability. However, the manufacturing of the semiconductors and hardware that enable AI also carries severe environmental impacts that receive comparatively less public attention. Building and operating a leading-edge semiconductor fabrication plant, or "fab", has substantial resource requirements and polluting byproducts beyond just a large carbon footprint. + +For example, a state-of-the-art fab producing state of the art chips like in 5nm can require up to [four million gallons of pure water each day](https://wccftech.com/tsmc-using-water-tankers-for-chip-production-as-5nm-plant-faces-rationing/). This water usage approaches what a city of half a million people would require for all needs. Sourcing this consistently places immense strain on local water tables and reservoirs, especially in already water-stressed regions which host many high-tech manufacturing hubs. + +Additionally, over 250 unique hazardous chemicals are utilized at various stages of semiconductor production within fabs [@mills1997overview]. These include volatile solvents like sulfuric acid, nitric acid, hydrogen fluoride, along with arsine, phosphine and other highly toxic substances. Preventing discharge of these chemicals requires extensive safety controls and wastewater treatment infrastructure to avoid soil contamination and risks to surrounding communities. Any improper chemical handling or unanticipated spill carries dire consequences. + +Beyond water consumption and chemical risks, fab operation also depends on rare metals sourcing, generates tons of dangerous waste products, and can hamper local biodiversity. This section will analyze these critical but less discussed impacts. With vigilance and investment in safety, the harms from semiconductor manufacturing can be contained while still enabling technological progress. However, ignoring these externalized issues will exacerbate ecological damage and health risks over the long run. + +### Water Usage and Stress {#water-usage-and-stress} + +Semiconductor fabrication is an incredibly water-intensive process. Based on an article from 2009, a typical 300mm silicon wafer requires 8,328 litres of water in total, of which 5,678 litres is ultrapure water [@cope2009pure]. Today, a typical fab can use up to [four million gallons of pure water](https://wccftech.com/tsmc-arizona-foundry-205-million-approved/). TSMC's latest fab in Arizona is projected to use 8.9 million gallons per day, or nearly 3 percent of the city's current water production, just to operate one facility. To put things in perspective, an by Intel and [Quantis](https://quantis.com/) found that over 97% of their direct water consumption is attributed to semiconductor manufacturing operations within their own fabrication facilities [@cooper2011semiconductor]. + +This water is used to flush away contaminants in cleaning steps repeatedly and also acts as a coolant and carrier fluid in thermal oxidation, chemical deposition, and chemical mechanical planarization processes. This approximates the daily water consumption of a city with a population of half a million people during peak summer months. + +Despite being located in regions with sufficient water, the intensive usage can severely depress local water tables and drainage basins. For example, the city of Hsinchu in Taiwan suffered [sinking water tables and seawater intrusion](https://wccftech.com/tsmc-using-water-tankers-for-chip-production-as-5nm-plant-faces-rationing/) into aquifers due to excessive pumping to satisfy water supply demands from the Taiwan Semiconductor Manufacturing Company (TSMC) fab. In water-scarce inland areas like Arizona, [massive water inputs are needed](https://www.americanbar.org/groups/environment_energy_resources/publications/wr/a-tale-of-two-shortages/) to support fabs despite already strained reservoirs. + +Besides depletion, water discharge from fabs also risks environmental contamination if not properly treated. While much discharge is recycled within the fab, the purification systems still filter out metals, acids, and other contaminants that can pollute rivers and lakes if not cautiously handled [@Prakash_2023]. These factors make managing water usage an essential consideration when mitigating wider sustainability impacts. + +### Hazardous Chemicals Usage {#hazardous-chemicals-usage} + +Modern semiconductor fabrication involves working with many highly hazardous chemicals under extreme conditions of heat and pressure [@kimChemicalUseSemiconductor2018]. Key chemicals utilized include: + +* **Strong acids:** Hydrofluoric, sulfuric, nitric, and hydrochloric acids rapidly eat through oxides and other surface contaminants but also pose toxicity dangers. Fabs can use thousands of metric tons of these acids annually. Accidental exposure can be fatal for workers. +* **Solvents:** Key solvents like xylene, methanol, methyl isobutyl ketone (MIBK) handle dissolving photoresists but have adverse health impacts like skin/eye irritation, narcotic effects if mishandled. They also create explosive and air pollution risks. +* **Toxic gases:** Gas mixtures containing arsine (AsH3), phosphine (PH3), diborane (B2H6), germane (GeH4), etc. are some of the deadliest chemicals used in doping and vapor deposition steps. Minimal exposures can lead to poisoning, tissue damage, and even death without quick treatment. +* **Chlorinated compounds:** Older chemical mechanical planarization formulations incorporated perchloroethylene, trichloroethylene and other chlorinated solvents since banned due to carcinogenic effects and ozone layer impacts. However, their prior release still threatens surrounding groundwater sources. + +Strict handling protocols, protective equipment for workers, ventilation, filtrating/scrubbing systems, secondary containment tanks, and specialized disposal mechanisms are vital where these chemicals are used to minimize health, explosion, air, and environmental spill dangers [@waldSemiconductorManufacturingIntroduction1987]. But human errors and equipment failures still occasionally occur--highlighting why reducing fab chemical intensities is an ongoing sustainability effort. + +### Resource Depletion {#resource-depletion} + +While silicon forms the base, there is an almost endless supply of silicon available on Earth. In fact, [silicon is the second most plentiful element found in the Earth's crust](https://en.wikipedia.org/wiki/Abundance_of_elements_in_Earth%27s_crust), accounting for 27.7% of the crust's total mass. Only oxygen exceeds silicon in abundance within the crust. Therefore, silicon is not necessary to consider for resource depletion. However, the various specialty metals and materials that enable the integrated circuit fabrication process and provide specific properties are scarce. Maintaining supplies of these resources is crucial yet threatened by finite availability and geopolitical influences [@nakano2021geopolitics]. + +Gallium, indium, and arsenic are vital ingredients in forming ultra-efficient compound semiconductors used in highest speed chips suited for 5G and AI applications [@chenGalliumIndiumArsenic2006]. However, these rare elements have relatively scarce natural deposits that are being depleted. The United States Geological Survey has indium on its list of most critical at-risk commodities--estimated to have less than a 15 year viable global supply at current demand growth [@daviesEndangeredElements2011]. + +Helium is required in huge volumes for next-gen fabs to enable precise wafer cooling during operation. But helium’s relative rarity and the fact that once it vents into the atmosphere it quickly escapes Earth makes maintaining helium supplies extremely challenging long-term [@daviesEndangeredElements2011]. Substantial price increases and supply shocks are already occurring in this thinly-traded market according to the US National Academies. + +Other risks include how China controls over 90% of the rare earth elements critical to semiconductor materials production [@jhaRareEarthMaterials2014]. Any supply chain issues or trade disputes can lead to catastrophic raw material shortages given lack of current alternatives. In conjunction with helium shortages, resolving the limited availability and geographic imbalance in accessing essential ingredients remains a sector priority for sustainability. + +### Hazardous Waste Generation {#hazardous-waste-generation} + +Semiconductor fabs generate tons of hazardous waste annually as byproducts from the various chemical processes involved [@grossmanHighTechTrash2007]. The key waste streams include: + +* **Gaseous waste:** Fab ventilation systems capture harmful gases like arsine, phosphine, germane and filter them out to avoid worker exposure. But this produces significant quantities of dangerous condensed gas in need of specialized treatment. +* **VOCs:** Volatile organic compounds like xylene, acetone, methanol are used extensively as photoresist solvents and get evaporated as emissions during baking, etching, and stripping stages. VOCs pose toxicity issues and require scrubbing systems to prevent release. +* **Spent acids:** Strong acids such as sulfuric acid, hydrofluoric acid, nitric acid get depleted in cleaning and etching steps transforming into a corrosive toxic soup that can dangerously react releasing heat and fumes if mixed. +* **Sludge:** Water treatment of discharged effluent contains concentrated heavy metals, acid residues, and chemical contaminants. Filter press systems separate this hazardous sludge. +* **Filter cake:** Gaseous filtration systems generate multi-ton sticky cakes of dangerous absorbed compounds requiring containment. + +Without proper handling procedures, storage tanks, packaging materials, and secondary containment--improper disposal of any of these waste streams can lead to dangerous spills, explosions, and environmental release. And the massive volumes mean even well-run fabs produce tons of hazardous waste year after year requiring extensive treatment. + +### Biodiversity Impacts {#biodiversity-impacts} + +#### Habitat Disruption and Fragmentation {#habitat-disruption-and-fragmentation} + +Semiconductor fabs require large, contiguous land areas to accommodate cleanrooms, support facilities, chemical storage, waste treatment, and ancillary infrastructure. Developing these vast built-up spaces inevitably dismantles existing habitats, damaging sensitive biomes that may have taken decades to develop. For example, constructing a new fabrication module may level local forest ecosystems relied upon by species like spotted owls and elk for survival. The outright removal of such habitats severely threatens any wildlife populations dependant on those lands. + +Furthermore, the pipelines, water channels, air and waste exhaust systems, access roads, transmission towers and other support infrastructure fragments the remaining undisturbed habitats. Animals ranging in their daily movements for food, water and spawning can find migration patterns blocked by these physical human barriers bisecting previously natural corridors. + +#### Aquatic Life Disturbances {#aquatic-life-disturbances} + +With semi-conductor fabs consuming millions of gallons of ultra-pure water daily, accessing and discharging such volumes risks altering the suitability of nearby aquatic environments housing fish, water plants, amphibians and other species. If the fab is tapping groundwater tables as its primary supply source, overdrawing at unsustainable rates can deplete lakes or lead to drying of streams as water levels drop [@daviesEndangeredElements2011]. + +Additionally, discharging higher temperature wastewater used for cooling fabrication equipment can shift downstream river conditions through thermal pollution. Temperature changes beyond thresholds which native species evolved for can disrupt reproductive cycles. Warmer water also holds less dissolved oxygen critical to support aquatic plant and animal life [@poff2002aquatic]. Combined with traces of residual contaminants that escape filtration systems, the discharged water can cumulatively transform environments to be far less habitable for sensitive organisms [@tillFishDieoffsAre2019]. + +#### Air and Chemical Emissions {#air-and-chemical-emissions} + +While modern semiconductor fabs aim to contain air and chemical discharges through extensive filtration systems, some level of emissions often persist raising risks for nearby flora and fauna. Air pollutants including volatile organic compounds (VOCs), nitrogen oxide compounds (NOxs), and particulate matter from fab operational exhausts as well as power plant fuel emissions can carry downwind. + +As contaminants permeate local soils and water sources, wildlife ingesting affected food and water ingest toxic substances which research shows can hamper cell function, reproduction rates and longevity--slowly poisoning ecosystems [@hsu2016accumulation]. + +Likewise, accidental chemical spills and improper waste handling which releases acids, BODs, and heavy metals into soils can dramatically affect retention and leeching capabilities. Flora such as vulnerable native orchids adapted to nutrient-poor substrates can experience die-offs when contacted by foreign runoff chemicals that alter soil pH and permeability. One analysis found that a single 500 gallon nitric acid spill led to the regional extinction of a rare moss species in the year following when the acidic effluent reached nearby forest habitats. Such contamination events set off chain reactions across the interconnected web of life. Thus strict protocols are essential to avoid hazardous discharge and runoff. + +## Life Cycle Analysis {#life-cycle-analysis} + +Understanding the holistic environmental impact of AI systems requires a comprehensive approach that considers the entire life cycle of these technologies. Life Cycle Analysis (LCA) refers to a methodological framework used to quantify the environmental impacts across all stages in the lifespan of a product or system, from raw material extraction to end-of-life disposal. Applying LCA to AI systems can help identify priority areas to target for reducing overall environmental footprints. + +### Stages of an AI System's Life Cycle + +The life cycle of an AI system can be divided into four key phases as shown in @fig-lifecycle: + +* **Design Phase:**This includes the energy and resources used in the research and development of AI technologies. It encompasses the computational resources used for algorithm development and testing contributing to carbon emissions. + +* **Manufacture Phase:**This stage involves producing hardware components such as graphics cards, processors, and other computing devices necessary for running AI algorithms. Manufacturing these components often involves significant energy use for material extraction, processing, and greenhouse gas emissions. + +* **Use Phase:**The next most energy-intensive phase involves the operational use of AI systems. It includes the electricity consumed in data centers for training and running neural networks and powering end-user applications. This is arguably one of the most carbon-intensive stages. + +* **Disposal Phase:**This final stage covers the end-of-life aspects of AI systems, including the recycling and disposal of electronic waste generated from outdated or non-functional hardware past their usable lifespan. + +![A chart depicting the stages of the AI lifecycle, broken into 3 parts including Design, Develop, and Deploy.](images/sustainable_ai/ai_lifecycle.jpeg){#fig-lifecycle} + +### Environmental Impact at Each Stage {#environmental-impact-at-each-stage} + +**Design and Manufacturing** + +The environmental impact during these beginning-of-life phases includes emissions from energy use and resource depletion from extracting materials for hardware production. At the heart of AI hardware are semiconductors, primarily silicon, used to make the integrated circuits in processors and memory chips. This hardware manufacturing relies on metals like copper for wiring, aluminum for casings, and various plastics and composites for other components. It also uses rare earth metals and specialized alloys--elements like neodymium, terbium, and yttrium, are used in small but vital quantities. For example, the creation of GPUs relies on copper and aluminum. At the same time, chips use rare earth metals--the mining process for which can generate substantial carbon emissions and ecosystem damage. + +**Use Phase** + +AI computes the majority of emissions in the lifecycle due to continuous high-power consumption, especially for training and running models. This includes direct emissions from electricity usage and indirect emissions from non-renewable grid energy generation. Studies estimate training complex models can have a carbon footprint comparable to the lifetime emissions of up to five cars. + +**Disposal Phase** + +The impact of the disposal stage includes air and water pollution from toxic materials in devices, challenges associated with complex electronics recycling, and contamination when improperly handled. Harmful compounds from burned e-waste are released into the atmosphere. At the same time, landfill leakage of lead, mercury and other materials poses risks of soil and groundwater contamination if not properly controlled. Implementing effective electronics recycling is crucial. + +## Challenges in LCA {#challenges-in-lca} + +### Lack of Consistency and Standards {#lack-of-consistency-and-standards} + +One major challenge facing life cycle analysis (LCA) for AI systems is the current lack of consistent methodological standards and frameworks. Unlike product categories like building materials that have developed international standards for LCA through ISO 14040, there are no firmly established guidelines tailored to analyzing the environmental footprint of complex information technology like AI. + +This absence of uniformity means researchers make differing assumptions and varying methodological choices. For example, a 2021 study from the University of Massachusetts Amherst [@strubellEnergyPolicyConsiderations2019] analyzed the life cycle emissions of several natural language processing models but only considered computational resource usage for training and omitted hardware manufacturing impacts. A more comprehensive 2020 study from Stanford University researchers [@strubellEnergyPolicyConsiderations2019] included emissions estimates from the production of relevant servers, processors, and other components, following an ISO-aligned LCA standard for computer hardware. However, these diverging choices in system boundaries and accounting approaches reduce robustness and prevent apples-to-apples comparisons of results. + +Having standardized frameworks and protocols tailored to the unique aspects and rapid update cycles of AI systems would provide more coherence. This could better equip researchers and developers to understand environmental hotspots, compare technology options, and accurately track progress on sustainability initiatives across the AI field. Industry groups and international standards bodies like the IEEE or ACM should prioritize addressing this methodological gap. + +### Data Gaps {#data-gaps} + +Another key challenge for comprehensive life cycle assessment of AI systems is substantial data gaps, especially regarding upstream supply chain impacts and downstream electronic waste flows. Most existing studies focus narrowly on the learner or usage phase emissions from computational power demands, which misses a significant portion of lifetime emissions [@guptaACTDesigningSustainable2022]. + +For example, little public data from companies exists quantifying energy use and emissions from manufacturing the specialized hardware components that enable AI--including high-end GPUs, ASIC chips, solid-state drives and more. Researchers often rely on secondary sources or generic industry averages to approximate production impacts. Similarly, there is limited transparency into downstream fate once AI systems are discarded after 4-5 years of usable lifespans on average. + +While electronic waste generation levels can be estimated, specifics on hazardous material leakage, recycling rates, and disposal methods for the complex components are hugely uncertain without better corporate documentation or regulatory reporting requirements. + +Even for the usage phase, the lack of fine-grained data on computational resource consumption for training different model types makes reliable per-parameter or per-query emissions calculations difficult. Attempts to create lifecycle inventories estimating average energy needs for key AI tasks exist [@henderson2020towards; @anthony2020carbontracker] but variability across hardware setups, algorithms, and input data uncertainty remains extremely high. + +The challenge is that tools like [CodeCarbon](https://codecarbon.io/) and [ML $\textrm{CO}_2$](https://mlco2.github.io/impact/#compute) but these are ad hoc approaches at best. Bridging the real data gaps with more rigorous corporate sustainability disclosures and mandated environmental impact reporting will be key for AI’s overall climatic impacts to be understood and managed. + +### Rapid Pace of Evolution {#rapid-pace-of-evolution} + +The extremely quick evolution of AI systems poses additional challenges when it comes to keeping life cycle assessments up-to-date and accounting for the latest hardware and software advancements. The core algorithms, specialized chips, frameworks, and technical infrastructure underpinning AI have all been advancing at exceptionally fast rates, with new developments rapidly rendering prior systems obsolete. + +For example, in the deep learning space, novel neural network architectures that achieve significantly better performance on key benchmarks or new optimized hardware like Google's TPU chips can completely change what an “average” model looks like in less than a year. These swift shifts make one-off LCA studies outdated quickly for accurately tracking emissions from designing, running, or disposing of the latest AI. + +However, the resources and access required to continuously update LCAs also poses barriers. Frequently re-doing labor and data intensive life cycle inventories and impact modeling to stay current with AI’s state of the art is likely infeasible for many researchers and organizations. But without updated analyses, the environmental hotspots as algorithms and silicon chips continue rapidly evolving could be missed. + +This presents a difficulty in balancing dynamic precision through continuous assessment with pragmatic constraints. Some researchers have proposed simplified proxy metrics like tracking hardware generations over time or using representative benchmarks as an oscillating set of goalposts for relative comparisons, though granularity may be sacrificed. Overall, the challenge of rapid change will require innovative methodological solutions to prevent underestimating AI’s evolving environmental burdens. + +### Supply Chain Complexity {#supply-chain-complexity} + +Finally, the complex and often opaque supply chains associated with producing the wide array of specialized hardware components that enable AI pose challenges for comprehensive life cycle modeling. State-of-the-art AI relies on leveraging cutting-edge advancements in processing chips, graphics cards, data storage, networking equipment and more. However, tracking emissions and resource use across the tiered networks of globalized suppliers for all these components is extremely difficult. + +For example, NVIDIA graphics processing units dominate much AI computing hardware, but the company relies on over several discrete suppliers across Asia and beyond to produce the GPUs. Many firms at each supplier tier choose not to disclose facility-level environmental data that could enable robust LCAs fully. Gaining end-to-end transparency down multiple levels of suppliers across disparate geographies with varying disclosure protocols and regulations poses barriers, despite being crucial for complete boundary setting. This becomes even more complex when attempting to model emerging hardware accelerators like tensor processing units (TPUs), whose production networks still need to be made public. + +Without willingness from tech giants to require and consolidate environmental impact data disclosure from across their global electronics supply chains, considerable uncertainty will remain around quantifying the full lifecycle footprint of AI hardware enablement. More supply chain visibility coupled with standardized sustainability reporting frameworks specifically addressing AI’s complex inputs hold promise for enriching LCAs and prioritizing environmental impact reductions. + +## Sustainable Design and Development {#sustainable-design-and-development} + +### Sustainability Principles {#sustainability-principles} + +As the impact of AI on the environment becomes increasingly evident, the focus on sustainable design and development in AI is gaining prominence. This involves incorporating sustainability principles into AI design, developing energy-efficient models, and integrating these considerations throughout the AI development pipeline. There is a growing need to consider its sustainability implications and develop principles to guide responsible innovation. Below is a core set of principles. The principles flows from the conceptual foundation, to practical execution, to supporting implementation factors, the principles provide a full cycle perspective on embedding sustainability in AI design and development. + +**Lifecycle Thinking:**Encouraging designers to consider the entire lifecycle of AI systems, from data collection and preprocessing to model development, training, deployment, and monitoring. The goal is to ensure sustainability is considered at each stage. This includes using energy-efficient hardware, prioritizing renewable energy sources, and planning to reuse or recycle retired models. + +**Future Proofing:**Designing AI systems anticipating future needs and changes can enhance sustainability. This may involve making models adaptable via transfer learning and modular architectures. It also includes planning capacity for projected increases in operational scale and data volumes. + +**Efficiency and Minimalism:**This principle focuses on creating AI models that achieve desired results with the least possible resource use. It involves simplifying models and algorithms to reduce computational requirements. Specific techniques include pruning redundant parameters, quantizing and compressing models, and designing efficient model architectures, such as those discussed in the [Optimizations](./optimizations.qmd) chapter. + +**Lifecycle Assessment (LCA) Integration:** Analyzing environmental impacts throughout the development and deployment lifecycles highlights unsustainable practices early on. Teams can then make needed adjustments, instead of discovering issues late when they are more difficult to address. Integrating this analysis into the standard design flow avoids creating legacy sustainability problems. + +**Incentive Alignment:**Economic and policy incentives should promote and reward sustainable AI development. This may include government grants, corporate initiatives, industry standards, and academic mandates for sustainability. Aligned incentives enable sustainability to become embedded in AI culture. + +**Sustainability Metrics and Goals:** Metrics that measure sustainability factors like carbon usage and energy efficiency are important to establish clearly. Establishing clear targets for these metrics provides concrete guidelines for teams to develop responsible AI systems. Tracking performance on metrics over time shows progress towards set sustainability goals. + +**Fairness, Transparency, and Accountability:**Sustainable AI systems should be fair, transparent, and accountable. Models should be unbiased, with transparent development processes and mechanisms for auditing and redressing issues. This builds public trust and enables the identification of unsustainable practices. + +**Cross-disciplinary Collaboration:**AI researchers teaming up with environmental scientists and engineers can lead to innovative systems that are high-performing yet environmentally friendly. Combining expertise from different fields from the start of projects enables sustainable thinking to be incorporated into the AI design process. + +**Education and Awareness:** Workshops, training programs, and course curricula that cover AI sustainability raise awareness among the next generation of practitioners. This equips students with the knowledge to develop AI that consciously minimizes negative societal and environmental impacts. Instilling these values from the start shapes tomorrow's professionals and company cultures. + +## Green AI Infrastructure {#green-ai-infrastructure} + +Green AI represents a transformative approach to AI that incorporates environmental sustainability as a fundamental principle across the AI system design and lifecycle [@schwartz2020green]. This shift is driven by growing awareness of AI technologies' significant carbon footprint and ecological impact, especially the compute-intensive process of training complex ML models. + +The essence of Green AI lies in its commitment to align AI advancement with sustainability goals around energy efficiency, renewable energy usage, and waste reduction. The introduction of Green AI ideals reflects maturing responsibility across the tech industry towards environmental stewardship and ethical technology practices. It moves beyond technical optimizations towards holistic life cycle assessment on how AI systems affect sustainability metrics. Setting new bars for ecologically conscious AI paves the way for the harmonious coexistence of technological progress and planetary health. + +### Energy Efficient AI Systems {#energy-efficient-ai-systems} + +Energy efficiency in AI systems is a cornerstone of Green AI, aiming to reduce the significant energy demands traditionally associated with AI development and operations. This shift towards energy-conscious AI practices is vital in addressing the environmental concerns raised by the rapidly expanding field of AI. By focusing on energy efficiency, AI systems can become more sustainable, lessening their environmental impact and paving the way for more responsible AI use. + +As we have discussed earlier, the training and operation of AI models, especially large-scale ones, are known for their high energy consumption stemming from compute-intensive model architecture and reliance on vast amounts of training data. For example, it is estimated that training a large state-of-the-art neural network model can have a carbon footprint of 284 tonnes--equivalent to the lifetime emissions of 5 cars [@strubellEnergyPolicyConsiderations2019]. + +To tackle the massive energy demands, researchers and developers are actively exploring methods to optimize AI systems for better energy efficiency without losing model accuracy or performance. This includes techniques like the ones we have discussed in the model optimizations, efficient AI and hardware acceleration chapters: + +* Knowledge distillation to transfer knowledge from large AI models to miniature versions +* Quantization and pruning approaches that reduce computational and space complexities +* Low-precision numerics--lowering mathematical precision without impacting model quality +* Specialized hardware like TPUs, neuromorphic chips tuned explicitly for efficient AI processing + +One example is Intel’s work on Q8BERT – quantizing BERT language model with 8-bit integers, leading to 4x reduction in model size with minimal accuracy loss [@zafrir2019q8bert]. The push for energy-efficient AI is not just a technical endeavor--it has tangible real-world implications. More performant systems lower AI's operational costs and carbon footprint, making it accessible for widespread deployment on mobile and edge devices. It also paves the path toward the democratization of AI and mitigates unfair biases that can emerge from uneven access to computing resources across regions and communities. Pursuing energy-efficient AI is thus crucial for creating an equitable and sustainable future with AI. + +### Sustainable AI Infrastructure {#sustainable-ai-infrastructure} + +Sustainable AI infrastructure includes the physical and technological frameworks that support AI systems, focusing on environmental sustainability. This involves designing and operating AI infrastructure in a way that minimizes ecological impact, conserves resources, and reduces carbon emissions. The goal is to create a sustainable ecosystem for AI that aligns with broader environmental objectives. + +Central to sustainable AI infrastructure are green data centers, which are optimized for energy efficiency and often powered by renewable energy sources. These data centers employ advanced cooling technologies [@ebrahimi_review_2014], energy-efficient server designs [@uddin_energy_2012], and smart management systems [@buyya2010energyefficient] to reduce power consumption. The shift towards green computing infrastructure also involves adopting energy-efficient hardware, like AI-optimized processors that deliver high performance with lower energy requirements, which we discussed in the [AI Acceleration](./hw_acceleration.qmd) chapter. These efforts collectively reduce the carbon footprint of running large-scale AI operations. + +Integrating renewable energy sources, such as solar, wind, and hydroelectric power, into AI infrastructure is important for environmental sustainability [@chua1971memristor]. Many tech companies and research institutions are [investing in renewable energy projects to power their data centers](https://www.forbes.com/sites/siemens-smart-infrastructure/2023/03/13/how-data-centers-are-driving-the-renewable-energy-transition/?sh=3208c5b54214). This not only helps in making AI operations carbon-neutral but also promotes the wider adoption of clean energy. Using renewable energy sources is a clear statement of commitment to environmental responsibility in the AI industry. + +Sustainability in AI also extends to the materials and hardware used in creating AI systems. This involves choosing environmentally friendly materials, adopting recycling practices, and ensuring responsible electronic waste disposal. Efforts are underway to develop more sustainable hardware components, including energy-efficient chips designed for domain-specific tasks (such as AI accelerators) and environmentally friendly materials in device manufacturing [@cenci_eco-friendly_2022;@irimia-vladu_green_2014]. The lifecycle of these components is also a focus, with initiatives aimed at extending the lifespan of hardware and promoting recycling and reuse. + +While strides are being made in sustainable AI infrastructure, challenges remain, such as the high costs of green technology and the need for global standards in sustainable practices. Future directions may include more widespread adoption of green energy, further innovations in energy-efficient hardware, and international collaboration on sustainable AI policies. The pursuit of sustainable AI infrastructure is not just a technical endeavor but a holistic approach that encompasses environmental, economic, and social aspects, ensuring that AI advances in harmony with our planet's health. + +### Frameworks and Tools {#frameworks-and-tools} + +To effectively implement Green AI practices, it is essential to have access to the right frameworks and tools. These resources are designed to assist developers and researchers in creating more energy-efficient and environmentally friendly AI systems. They range from software libraries optimized for low-power consumption to platforms that facilitate the development of sustainable AI applications. + +There are several software libraries and development environments specifically tailored for Green AI. These tools often include features for optimizing AI models to reduce their computational load and, consequently, their energy consumption. For example, libraries in PyTorch and TensorFlow that support model pruning, quantization, and efficient neural network architectures enable developers to build AI systems that require less processing power and energy. Additionally, there are open source communities like the [Green Carbon Foundation](https://github.com/Green-Software-Foundation) creating a centralized carbon intensity metric and building software for carbon-aware computing. + +Energy monitoring tools are crucial for Green AI, as they allow developers to measure and analyze the energy consumption of their AI systems. By providing detailed insights into where and how energy is being used, such as in @fig-azuredashboard, these tools enable developers to make informed decisions about optimizing their models for better energy efficiency. This can involve adjustments in algorithm design, hardware selection, cloud computing software selection, or operational parameters. + +![Microsoft Azure now provides [dashboards](https://techcommunity.microsoft.com/t5/green-tech-blog/charting-the-path-towards-sustainable-ai-with-azure-machine/ba-p/2866923) to describe energy consumption for a job that is run](images/sustainable_ai/azure_dashboard.png){#fig-azuredashboard} + +With the increasing integration of renewable energy sources in AI operations, frameworks that facilitate this process are becoming more important. These frameworks help manage the energy supply from renewable sources like solar or wind power, ensuring that AI systems can operate efficiently with fluctuating energy inputs. + +Beyond energy efficiency, sustainability assessment tools help evaluate the broader environmental impact of AI systems. These tools can analyze factors like the carbon footprint of AI operations, the lifecycle impact of hardware components [@guptaACTDesigningSustainable2022], and the overall sustainability of AI projects [@prakashCFUPlaygroundFullStack2023]. + +The availability and ongoing development of Green AI frameworks and tools are critical for advancing sustainable AI practices. By providing the necessary resources for developers and researchers, these tools facilitate the creation of more environmentally friendly AI systems and encourage a broader shift towards sustainability in the tech community. As Green AI continues to evolve, these frameworks and tools will play a vital role in shaping a more sustainable future for AI. + +## Case Study: Google’s 4Ms {#case-study-google-4ms} + +Over the past decade, AI has rapidly moved from the realm of academic research to large-scale production systems powering numerous Google products and services. As AI models and workloads have grown exponentially in size and computational demands, concerns have emerged about their energy consumption and carbon footprint. Some researchers predicted runaway growth in ML's energy appetite that could outweigh efficiencies gained from improved algorithms and hardware [@9563954]. + +However, Google's own production data reveals a different story--with AI representing a steady 10-15% of total company energy usage from 2019 to 2021. This case study analyzes how Google applied a systematic approach leveraging four best practices--what they term the "4 Ms" of model efficiency, machine optimization, mechanization through cloud computing, and mapping to green locations to bend the curve on emissions from AI workloads. + +The scale of Google's AI usage makes it an ideal case study. In 2021 alone, the company was training models like the 1.2 trillion parameter GLam model. Analyzing how the application of AI has been paired with rapid efficiency gains in this environment helps us by providing a logical blueprint for the broader AI field to follow. + +By transparently publishing detailed energy usage statistics, adoption rates of carbon-free clouds and renewables purchases, and more alongside its technical innovations, Google has enabled outside researchers to accurately measure progress. Their study in the ACM CACM [@patterson2022carbon] highlights how the company's multi-pronged approach shows that predictions of runaway AI energy consumption can be overcome through focusing engineering efforts on sustainable development patterns. The pace of improvements also suggests ML's efficiency gains are just getting started. + +### Google’s 4M Best Practices {#google-4m-best-practices} + +To curb emissions from their rapidly expanding AI workloads, Google engineers systematically identified four best practice areas--termed the “4 Ms”--where optimizations could compound to reduce the carbon footprint of ML: + +* Model - Selecting efficient AI model architectures can reduce computation by 5-10X with no loss in model quality. Google has focused extensive research on developing sparse models and neural architecture search to create more efficient models like the Evolved Transformer and Primer. +* Machine - Using hardware optimized for AI over general purpose systems improves performance per watt by 2-5X. Google's Tensor Processing Units (TPUs) led to 5-13X better carbon efficiency versus GPUs not optimized for ML. +* Mechanization - By leveraging cloud computing systems tailored for high utilization over conventional on-premise data centers, energy costs reduce by 1.4-2X. Google cites its data centers' Power Usage Effectiveness outpacing industry averages. +* Map - Choosing data center locations with low-carbon electricity reduces gross emissions by another 5-10X. Google provides real-time maps highlighting its renewable energy percentage by facility. + +Together, these practices created drastic compound efficiency gains. For example, optimizing the Transformer AI model on TPUs in a sustainable data center location cut energy use by a factor of 83 and lowered $\textrm{CO}_2$ emissions by a factor of 747. + +### Significant Results {#significant-results} + +Google's efforts to improve the carbon efficiency of ML have produced measurable gains helping to restrain overall energy appetite, despite exponential growth in AI adoption across products and services. One key datapoint highlighting this progress is that AI workloads have remained a steady 10% to 15% of total company energy use from 2019 to 2021. As AI became integral to ever more Google offerings, overall compute cycles dedicated to AI grew substantially. However, efficiencies on algorithms, specialized hardware, data center design and flexible geography allowed sustainability to keep pace – with AI representing just a fraction of total data center electricity over years of expansion. + +Other case studies further underscore how an engineering focus on sustainable AI development patterns enabled rapid quality improvements in lockstep with environmental gains. For example, the natural language processing model GPT-3 was viewed as state-of-the-art in mid-2020. Yet its successor GLaM improved accuracy while cutting training compute needs and using cleaner data center energy--cutting CO2 emissions by a factor of 14 in just 18 months of model evolution. + +Similarly, Google found past published speculation missing the mark on ML’s energy appetite by factors of 100 to 100,000X due to lacking real-world metrics. By transparently tracking optimization impact, Google hoped to motivate efficiency while preventing overestimated extrapolations about ML’s environmental toll. + +Together these data-driven case studies show how companies like Google are steering AI advancements toward sustainable trajectories and driving efficiency improvements to outpace adoption growth. And with further efforts around lifecycle analysis, inference optimization, and renewable expansion, companies can aim to accelerate progress – giving evidence that ML’s clean potential is only just being unlocked by current gains. + +### Further Improvements {#further-improvements} + +While Google has made measurable progress in restraining the carbon footprint of its AI operations, the company recognizes further efficiency gains will be vital for responsible innovation given the technology’s ongoing expansion. + +One area of focus is showing how advances often incorrectly viewed as increasing unsustainable computing – like neural architecture search (NAS) to find optimized models – actually spur downstream savings outweighing their upfront costs. Despite expending more energy for model discovery rather than hand-engineering, NAS cuts lifetime emissions by producing efficient designs callable across countless applications. + +Additionally, analysis reveals focusing sustainability efforts on data center and server-side optimization makes sense given the dominant energy draw versus consumer devices. Though Google aims to shrink inference impacts across processors like mobile phones, priority rests on improving training cycles and data center renewables procurement for maximal effect. + +To that end, Google’s progress in pooling compute in efficiently designed cloud facilities highlights the value of scale and centralization. As more workloads shift away from inefficient on-premise servers, internet giants’ prioritization of renewable energy – with Google and Facebook matched 100% by renewables since 2017 and 2020 respectively – unlocks compounding emissions cuts. + +Together these efforts emphasize that while no resting on laurels is possible, Google’s multipronged approach shows AI efficiency improvements are only accelerating. Cross-domain initiatives around lifecycle assessment, carbon-conscious development patterns, transparency, and matching rising AI demand with clean electricity supply pave a path toward bending the curve further as adoption grows. The company’s results compel the broader field towards replicating these integrated sustainability pursuits. + +## Embedded AI - Internet of Trash {#embedded-ai-internet-of-trash} + +While much attention has focused on making the immense data centers powering AI more sustainable, an equally pressing concern is the movement of AI capabilities into smart edge devices and endpoints. Edge/embedded AI allows near real-time responsiveness without connectivity dependencies. It also reduces transmission bandwidth needs. However, the increase of tiny devices leads to other risks. + +Tiny computers, microcontrollers, and custom ASICs powering edge intelligence face size, cost and power limitations that rule out high-end GPUs used in data centers. Instead, they require optimized algorithms and extremely compact, energy-efficient circuitry to run smoothly. But engineering for these microscopic form factors opens up risks around planned obsolescence, disposability, and waste. @fig-iot-devices shows that the number of IoT devices is projected to [reach 30 billion connected devices by 2030](https://www.statista.com/statistics/1183457/iot-connected-devices-worldwide/). + +![Number of Internet of Things (IoT) connected devices worldwide from 2019 to 2023, with forecasts from 2022 to 2030 from [Statista](https://www.statista.com/statistics/1183457/iot-connected-devices-worldwide/).](images/sustainable_ai/statista_chip_growth.png){#fig-iot-devices} + +End-of-life handling of internet-connected gadgets embedded with sensors and AI remains an often overlooked issue during design, though these products permeate consumer goods, vehicles, public infrastructure, industrial equipment and more. + +#### E-waste {#e-waste} + +Electronic waste, or e-waste, refers to discarded electrical equipment and components that enter the waste stream. This includes devices that have to be plugged in, have a battery, or electrical circuitry. With the rising adoption of internet-connected smart devices and sensors, e-waste volumes are rapidly increasing each year. These proliferating gadgets contain toxic heavy metals like lead, mercury, and cadmium that become environmental and health hazards when improperly disposed. + +The amount of electronic waste being produced is growing at an alarming rate. Today, [we already produce 50 million tons per year](https://www.unep.org/news-and-stories/press-release/un-report-time-seize-opportunity-tackle-challenge-e-waste). By 2030, that figure is projected to jump to a staggering 75 million tons as consumer electronics consumption continues to accelerate. Global e-waste production is on track to reach 120 million tonnes per year by 2050 [@un_world_economic_forum_2019]. From smartphones and tablets to internet-connected devices and home appliances, the soaring production and short lifecycles of our gadgets is fueling this crisis. + +Developing nations are being hit the hardest as they lack the infrastructure to safely process obsolete electronics. In 2019, formal e-waste recycling rates in poorer countries ranged from just 13% to 23%. The remainder ends up illegally dumped, burned, or crudely dismantled--releasing toxic materials into the environment and harming workers as well as local communities. Clearly more needs to be done to build global capacity for ethical and sustainable e-waste management or we risk irreversible damage. + +The danger is that crude handling of electronics to strip valuables exposes marginalized workers and communities to noxious burnt plastics/metals. Lead poisoning poses especially high risks to child development if ingested or inhaled. Overall, only about 20% of e-waste produced was collected using environmentally sound methods according to UN estimates [@un_world_economic_forum_2019]. So solutions for responsible lifecycle management are urgently required to contain the unsafe disposal as volume soars higher. + +#### Disposable Electronics {#disposable-electronics} + +Rapidly falling costs of microcontrollers, tiny rechargeable batteries, and compact communication hardware has enabled embedding intelligent sensor systems throughout everyday consumer goods. These internet-of-things (IoT) devices monitor product conditions, user interactions, and environment factors in order to enable real-time responsiveness, personalization, and data-driven business decisions in the evolving connected marketplace. + +However, these embedded electronics face little oversight or planning around sustainably handling their eventual disposal once the often plastic-encased products get thrown out following brief lifetimes. IoT sensors now commonly reside in single-use items like water bottles, food packaging, prescription bottles, and cosmetic containers that overwhelmingly enter landfill waste streams after a few weeks to months of consumer use. + +The problem accelerates as more manufacturers rush to integrate mobile chips, power sources, Bluetooth modules and other modern silicon ICs costing under US$1 into various merchandise without protocols for recycling, replacing batteries or component reusability. Despite their small individual size, collectively the volumes of these devices and lifetime waste burden loom large. Unlike regulating larger electronics, few policy constraints currently exist around materials requirements or toxicity in tiny disposable gadgets. + +While offering convenience when working, the unsustainable combination of difficult retrievability and limited safe breakdown mechanisms causes disposable connected devices to contribute outsized shares of future e-waste volumes needing urgent attention. + +#### Planned Obsolescence {#planned-obsolescence} + +Planned obsolescence refers to the intentional design strategy of manufacturing products with artificially limited lifetimes that quickly become non-functional or outdated. This spurs faster replacement purchase cycles as consumers find devices no longer meeting needs within a few years. However, electronics designed for premature obsolescence contribute to unsustainable e-waste volumes. + +For example, gluing smartphone batteries and components together hinders repairability compared to using modular, accessible assemblies. Or rolling out software updates that deliberately slow system performance creates a perception worth upgrading devices produced only several years earlier. + +Likewise, fashionable introductions of new product generations with minor but exclusive feature additions makes prior versions rapidly seem dated. These tactics compel buying new gadgets ([e.g. Iphones](https://www.cnbc.com/2020/12/08/the-psychology-of-new-iphone-releases-apple-marketing.html)) long before operational endpoints. When multiplied across fast-paced electronics categories, the result is billions of barely worn items being discarded annually. + +Planned obsolescence thus intensifies resource utilization and waste creation in making products with no intention for long lifetimes. This contradicts sustainability principles around durability, reuse and material conservation. While stimulating continuous sales and gains for manufacturers in the short term, the strategy externalizes environmental costs and toxins onto communities lacking proper e-waste processing infrastructure. + +Policy and consumer action is crucial to counter gadget designs that are needlessly disposable by default. Companies should also invest in product stewardship programs supporting responsible reuse and reclamation. + +Consider the real world example. [Apple has faced scrutiny](https://undergradlawreview.blog.fordham.edu/consumer-protection/the-product-ecosystem-and-planned-obsolescence-apples-threats-to-consumer-rights/) over the years for allegedly engaging in planned obsolescence to encourage customers to buy new iPhone models. The company was allegedly designing its phones so that performance degrades over time or existing features become incompatible with new operating systems, which critics argue is meant to spur more rapid upgrade cycles. In 2020, Apple paid a 25 million Euros in fine to settle a case in France where regulators found the company guilty of intentionally slowing down older iPhones without clearly informing customers via iOS updates. + +By failing to be transparent about power management changes that reduced device performance, Apple participated in deceptive activities that reduced product lifespan to drive sales. The company claimed it was done to “smooth out” peaks that could cause older batteries to shut down suddenly. But this is an example that clearly highlights the legal risks around employing planned obsolescence and not properly disclosing when functionality changes impact device usability over time--even leading brands like Apple can run into trouble if perceived to be intentionally shortening product life cycles. + +## Policy and Regulatory Considerations {#policy-and-regulatory-considerations} + +### Measurement and Reporting Mandates {#measurement-and-reporting-mandates} + +One policy mechanism with increasing relevance for AI systems is measurement and reporting requirements regarding energy consumption and carbon emissions. Mandated metering, auditing, disclosures, and more rigorous methodologies aligned to sustainability metrics can help address information gaps hindering efficiency optimizations. + +On the simple end, national or regional policies may require companies above a certain size utilizing AI in their products or backend systems to report energy consumption or emissions associated with major AI workloads. Organizations like the Partnership on AI, IEEE, and NIST could help shape standardized methodologies. More complex proposals involve defining consistent ways to measure computational complexity, data center PUE, carbon intensity of energy supply, and efficiencies gained through AI-specific hardware. + +Reporting obligations for public sector users procuring AI services--such as through proposed legislation in Europe--could also increase transparency. However, regulators must balance the additional measurement burden such mandates place on organizations versus ongoing carbon reductions from ingraining sustainability-conscious development patterns. + +To be most constructive, any measurement and reporting policies should focus on enabling continuous refinement rather than simplistic restrictions or caps. As AI advancements unfold rapidly, nimble governance guardrails that embed sustainability considerations into normal evaluation metrics can motivate positive change. But overprescription risks constraining innovation if requirements grow outdated. By combining flexibility with appropriate transparency guardrails, AI efficiency policy aims to accelerate progress industry-wide. + +### Restriction Mechanisms {#restriction-mechanisms} + +In addition to reporting mandates, policymakers have several restriction mechanisms that could directly shape how AI systems are developed and deployed to curb emissions: + +Caps on Computing Emissions: The [European Commission’s proposed AI Act](https://digital-strategy.ec.europa.eu/en/policies/european-approach-artificial-intelligence) takes a horizontal approach that could allow setting economy-wide caps on the volume of computing power available for training AI models. Similar to emissions trading systems, caps aim to indirectly disincentivize extensive computing over sustainability. However, model quality could suffer absent pathways for procuring additional capacity. + +Conditioning Access to Public Resources: Some experts have proposed incentives like only allowing access to public datasets or computing power for developing fundamentally efficient models rather than extravagant architectures. For example, the [MLCommons benchmarking consortium](https://mlcommons.org/) founded by major tech firms could formally integrate efficiency into its standardized leaderboard metrics. However, conditioned access risks limiting innovation. + +Financial Mechanisms: Analogous to carbon taxes on polluting industries, fees applied per unit of AI-related compute consumption could discourage unnecessary model scaling while funding efficiency innovations. Tax credits could alternatively reward organizations pioneering more accurate but compact AI techniques. But financial tools require careful calibration between revenue generation, fairness, and not over-penalizing productive uses of AI. + +Technology Bans: If measurement consistently pinned extreme emissions on specific applications of AI without paths for remediation, outright bans present a tool of last resort for policymakers. However, given AI’s dual use, defining harmful versus beneficial deployments proves complex, necessitating holistic impact assessment before concluding no redeeming value exists. Banning promising technologies risks unintended consequences and requires caution. + +### Government Incentives {#government-incentives} + +It is a common practice for governments to provide tax or other incentives to consumers or businesses when contributing to more sustainable practices in technology. Such incentives already exist in the US for [adopting solar panels](https://www.irs.gov/credits-deductions/residential-clean-energy-credit) or [energy efficient buildings](https://www.energy.gov/eere/buildings/179d-commercial-buildings-energy-efficiency-tax-deduction). To the best of our knowledge, no such tax incentives exist for AI specific development practices yet. + +Another potential incentive program that is beginning to be explored is the use of government grants to fund Green AI projects. For example, in Spain, [300 million euros have been allocated](https://www.state.gov/artificial-intelligence-for-accelerating-progress-on-the-sustainable-development-goals-addressing-societys-greatest-challenges/) to specifically fund projects in AI and sustainability. Government incentives are a promising avenue to encourage sustainable practices in business and consumer behavior, but they require careful thought into how those incentives will fit into market demands [@maxime2016impact]. + +### Self-Regulation {#self-regulation} + +Complimentary to potential government action, voluntary self-governance mechanisms allow the AI community to pursue sustainability ends without top-down intervention: + +Renewables Commitments: Large AI practitioners like Google, Microsoft, Amazon and Facebook have pledged to procure enough renewable electricity to match 100% of their energy demands. These commitments unlock compounding emissions cuts as compute scales up. Formalizing such programs incentivizes green data center regions. However, there are critiques to whether these pledges are enough [@monyei2018electrons]. + +Internal Carbon Prices: Some organizations utilize shadow prices on carbon emissions to represent environmental costs in capital allocation decisions between AI projects. If modeled effectively, theoretical charges on development carbon footprints steer funding toward efficient innovations rather than solely accuracy gains. + +Efficiency Development Checklists: Groups like the AI Sustainability Coalition suggest voluntary checklist templates highlighting model design choices, hardware configurations, and other factors architects can tune per application to restrain emissions. By ingraining sustainability as a primary success metric alongside accuracy and cost, organizations can drive change. + +Independent Auditing: Even absent public disclosure mandates, firms specializing in technology sustainability audits help AI developers identify waste, create efficiency roadmaps, and benchmark progress via impartial reviews. Structuring such audits into internal governance procedures or the procurement process expands accountability. + +### Global Considerations {#global-considerations} + +While measurement, restrictions, incentives, and self-regulation all represent potential policy mechanisms for furthering AI sustainability, fragmentation across national regimes risks unintended consequences. As with other technology policy domains, divergence between regions must be carefully managed. + +For example, OpenAI barred access to its viral ChatGPT chatbot for European users over data privacy concerns in the region. This came after the EU’s proposed AI Act signaled a precautionary approach allowing the EC to ban certain AI uses deemed high-risk, enforcing transparency rules that create uncertainty for release of brand new models. However, it would be wise to caution regulator action as it could inadvertently limit European innovation if regimes with lighter touch regulation attract more private sector AI research spending and talent. Finding common ground is key. + +The OECD principles on AI and the United Nations frameworks underscore universally agreed tenets all national policies should uphold: transparency, accountability, bias mitigation, and more. Constructively embedding sustainability as a core principle for responsible AI within such international guidance can motivate unified action without sacrificing flexibility across divergent legal systems. Avoiding race-to-the-bottom dynamics hinges on enlightened multilateral cooperation. + +## Public Perception and Engagement {#public-perception-and-engagement} + +As societal attention and policy efforts aimed at environmental sustainability ramp up worldwide, there is growing enthusiasm around leveraging AI to help address ecological challenges. However, public understanding and attitudes towards the role of AI systems in sustainability contexts remain mixed and clouded by misconceptions. On one hand, people hope advanced algorithms can provide new solutions for green energy, responsible consumption, decarbonization pathways and ecosystem preservation. But on the other, fears regarding risks of uncontrolled AI also seep into the environmental domain and undermine constructive discourse. Furthermore, lack of public awareness on key issues like transparency in development of sustainability-focused AI tools as well as potential biases in data or modeling also threaten to limit inclusive participation and degrade public trust. + +Tackling complex, interdisciplinary priorities like environmental sustainability requires informed, nuanced public engagement along with responsible advances in AI innovation itself. The path forward demands careful, equitable collaborative efforts between experts in fields like ML, climate science, environmental policy, social science and communication. Mapping the landscape of public perceptions, identifying pitfalls, and charting strategies to cultivate understandable, accessible and trustworthy AI systems targeting shared ecological priorities will prove essential to realizing sustainability goals. This complex terrain warrants deep examination into the sociotechnical dynamics involved. + +### AI Awareness {#ai-awareness} + +In May 2022, [Pew Research Center polled 5,101 U.S. adults](https://www.pewresearch.org/internet/2023/08/17/what-americans-know-about-ai-cybersecurity-and-big-tech/) finding 60% had heard or read "a little" about AI while 27% heard "a lot"--indicating decent broad recognition, but likely limited comprehension about details or applications. However, among those with some AI familiarity, concerns emerge regarding risks of personal data misuse according to agreed terms. Still 62% felt AI could potentially ease modern life if applied responsibly. Yet specific understanding of sustainability contexts remains lacking. + +Studies attempting to categorize online discourse sentiments find a nearly even split between optimism and caution regarding deployment of AI for sustainability goals. Factors driving positivity include hopes around better forecasting of ecological shifts using ML models. Negativity arises from lack of confidence in self-supervised algorithms avoiding unintended consequences due to unpredictable human impacts on complex natural systems during training. + +The most prevalent public belief remains that while AI does harbor potential for accelerating solutions on issues like emission reductions and wildlife protections, inadequate safeguarding around data biases, ethical blindspots and privacy considerations pose underappreciated risks if pursued carelessly, especially at scale. This leads to hesitancy around unconditional support without evidence of deliberate, democratically guided development. + +### Messaging {#messaging} + +[Optimistic efforts](https://www.climatechange.ai/) are highlighting AI's sustainability promise emphasize potential for advanced ML to radically accelerate decarbonization effects from smart grids, personalized carbon tracking apps, automated building efficiency optimizations, and predictive analytics guiding targeted conservation efforts. More comprehensive real-time modeling of complex climate and ecological shifts using self-improving algorithms offers hope for mitigating biodiversity losses and averting worst case scenarios. + +However, [cautionary perspectives](https://time.com/6266923/ai-eliezer-yudkowsky-open-letter-not-enough/), such as the [Asilomar AI Principles](https://futureoflife.org/open-letter/ai-principles/), question whether AI itself could exacerbate sustainability challenges if improperly constrained. Rising energy demands of large scale computing systems and increasingly massive neural network model training conflicts with clean energy ambitions. Lack of diversity in data inputs or priorities of developers might inadvertently downplay urgent environmental justice considerations. Near term skeptical public engagement likely hinges on lack of perceivable safeguards against uncontrolled AI systems that are running amok on core ecological processes before our eyes. + +In essence, polarized framings either promote AI as an indispensable tool for sustainability problem-solving--if compassionately directed toward people and planet--or present AI as an amplifier of existing harms insidiously dominating hidden facets of natural systems central to all life. Overcoming such impasses demands balancing honest trade-off discussions with shared visions for equitable, democratically governed technological progress targeting restoration. + +### Equitable Participation {#equitable-participation} + +Ensuring equitable participation and access should form a cornerstone of any sustainability initiative with potential for major societal impacts. This principle applies equally to AI systems targeting environmental goals. However, commonly excluded voices like frontline, rural or indigenous communities and future generations not present to consent could suffer disproportionate consequences from technology transformations. For instance, the [Partnership on AI](https://partnershiponai.org) has launched events expressly targeting input from marginalized communities on deploying AI responsibly. + +Ensuring equitable access and participation should form a cornerstone of any sustainability initiative with potential for major societal impacts be it AI or otherwise. However, inclusive engagement on environmental AI relies partly on availability and understanding of fundamental computing resources. As the recent [OECD](https://www.oecd.org/) report on [National AI Compute Capacity](https://www.oecd.org/economy/a-blueprint-for-building-national-compute-capacity-for-artificial-intelligence-876367e3-en.htm) highlights [@//content/paper/876367e3-en], many countries currently lack data or strategic plans mapping needs for the infrastructure required to fuel AI systems. This policy blind-spot could constrain economic goals and exacerbate barriers to entry for marginalized populations. Their blueprint urges developing national AI compute capacity strategies along dimensions of capacity, accessibility, innovation pipelines and resilience to anchor innovation. Otherwise inadequacies in underlying data storage, model development platforms or specialized hardware could inadvertently concentrate AI progress in the hands of select groups. Therefore, planning for balanced expansion of fundamental AI computing resources via policy initiatives ties directly to hopes for democratized sustainability problem-solving using equitable and transparent ML tools. + +The key idea is that equitable participation in AI systems targeting environmental challenges relies in part on getting the underlying computing capacity and infrastructure right, which requires proactive policy planning from a national perspective. + +### Transparency {#transparency} + +As public sector agencies and private companies alike rush towards adopting AI tools to help tackle pressing environmental challenges, calls for transparency around the development and functionality of these systems has began to amplify. Explainable and interpretable ML features grow more crucial for building trust in emerging models aiming to guide consequential sustainability policies. Initiatives like the [Montreal Carbon Pledge](https://unfccc.int/news/montreal-carbon-pledge) brought tech leaders together to commit to publishing impact assessments before launching environmental systems, as pledged below: + +*"As institutional investors, we have a duty to act in the best long-term interests of our beneficiaries. In this fiduciary role, we believe that there are long-term investment risks associated with greenhouse gas emissions, climate change and carbon regulation. + +In order to better understand, quantify and manage the carbon and climate change related impacts, risks and opportunities in our investments, it is integral to measure our carbon footprint. Therefore, we commit, as a first step, to measure and disclose the carbon footprint of our investments annually with the aim of using this information to develop an engagement strategy and/or identify and set carbon footprint reduction targets."* + +We need a similar pledge for AI sustainability and responsibility. Widespread acceptance and impact of AI sustainability solutions will partly on deliberate communication of validation schemes, metrics, and layers of human judgment applied before live deployment. Efforts like [NIST’s Principles for Explainable AI](https://oecd.ai/en/dashboards/policy-initiatives/http:%2F%2Faipo.oecd.org%2F2021-data-policyInitiatives-26746) can be helpful for fostering transparency into AI systems. The National Institute of Standards and Technology (NIST) has published an influential set of guidelines dubbed the Principles for Explainable AI [@phillips2020four]. This framework articulates best practices for designing, evaluating and deploying responsible AI systems with transparent and interpretable features that build critical user understanding and trust. + +It delineates four core principles: Firstly, AI systems should provide contextually relevant explanations justifying the reasoning behind their outputs to appropriate stakeholders. Secondly, these AI explanations must communicate information in a truly meaningful way for their target audience’s appropriate comprehension level. Next, there is the accuracy principle which dictates explanations should faithfully reflect the actual process and logic informing an AI model’s internal mechanics for generating given outputs or recommendations based on inputs. Finally, a knowledge limits principle compels explanations to clarify an AI model's boundaries in capturing the full breadth of real-world complexity, variance and uncertainties within a problem space. + +Altogether, these NIST principles offer AI practitioners and adopters guidance on key transparency considerations vital for developing accessible solutions that prioritize user autonomy and trust rather than simply maximizing predictive accuracy metrics alone. As AI rapidly advances across sensitive social contexts like healthcare, finance, employment and beyond, such human centered design guidelines will continue growing in importance for anchoring innovation to public interests. + +This applies equally to the environmental ability domain. Overall, responsible and democratically guided AI innovation targeting shared ecological priorities depends on maintaining public vigilance, understanding, and oversight over otherwise opaque systems taking prominent roles in societal decisions. Prioritizing explainable algorithm designs and radical transparency practices per global standards can help sustain collective confidence that these tools improve rather than imperil hopes for AI driven future. + +## Future Directions and Challenges {#future-directions-and-challenges} + +As we look towards the future, the role of AI in environmental sustainability is poised to grow even more significant. The potential of AI to drive advancements in renewable energy, climate modeling, conservation efforts, and more is immense. However, it is a two-sided coin, as we need to overcome several challenges and direct our efforts towards sustainable and responsible AI development. + +### Future Directions {#future-directions} + +One of the key future directions is the development of more energy-efficient AI models and algorithms. This involves ongoing research and innovation in areas like model pruning, quantization, and the use of low-precision numerics, and developing the hardware to enable full profitability of these innovations. Even further, we look at alternative computing paradigms which do not rely on von-Neumann architectures. More on this topic can be found in the hardware acceleration chapter. The goal is to create AI systems that deliver high performance while minimizing energy consumption and carbon emissions. + +Another important direction is the integration of renewable energy sources into AI infrastructure. As data centers continue to be major contributors to AI's carbon footprint, transitioning to renewable energy sources like solar and wind is crucial. Developments in long-term, sustainable energy storage, such as [Ambri](https://ambri.com/), an MIT spinoff, could enable this transition. This requires significant investment and collaboration between tech companies, energy providers, and policymakers. + +### Challenges {#challenges} + +Despite these promising directions, several challenges need to be addressed. One of the major challenges is the lack of consistent standards and methodologies for measuring and reporting the environmental impact of AI. It is essential that the complexity of life cycles of both AI models and system hardware are captured by these methods. Next, efficient and environmentally-sustainable AI infrastructure and system hardware is needed. This consists of three components. Aimed at maximizing the utilization of accelerator and system resources, prolonging the lifetime of AI infrastructure, and designing systems hardware with environmental impact in mind. + +On the software side, we should make a trade-off between experimentation and the subsequent training cost. Techniques such as neural architecture search and hyperparameter optimization can be used for design space exploration. However, these are often very resource-intensive. Efficient experimentation can reduce the environmental footprint overhead significantly. Next, methods to reduce wasted training efforts should be explored. + +To improve model quality, we often scale the dataset. However, the increased system resources required for data storage and ingestion caused by this scaling has a significant environmental impact [@wu2022sustainable]. A thorough understanding of the rate at which data loses its predictive value and devising data sampling strategies is important. + +Data gaps also pose a significant challenge. Without companies and governments openly sharing detailed and accurate data on energy consumption, carbon emissions, and other environmental impacts, it is difficult to develop effective strategies for sustainable AI. + +Finally, the fast pace of AI development requires an agile approach to the policy imposed on these systems. The policy should ensure sustainable development without constraining innovation. This requires experts in all domains of AI, environmental sciences, energy and policy to work together to achieve a sustainable future. + +## Conclusion {#conclusion} + +​​As AI continues rapidly expanding across industries and society, we must address sustainability considerations. AI promises breakthrough innovations, yet its environmental footprint threatens its widespread growth. This chapter analyzes multiple facets, from energy and emissions to waste and biodiversity impacts, that AI/ML developers must weigh when creating responsible AI systems. + +Fundamentally, we require elevating sustainability as a primary design priority rather than an afterthought. Techniques like energy-efficient models, renewable-powered data centers, and hardware recycling programs offer solutions, but holistic commitment remains vital. We need standards around transparency, carbon accounting, and supply chain disclosures to supplement technical gains. Still, examples like Google’s 4M efficiency practices containing ML energy use highlight that with concerted effort, we can advance AI in lockstep with environmental objectives. We achieve this harmonious balance by having researchers, corporations, regulators and users collaborate across domains. The aim is not perfect solutions but rather continuous improvement as we integrate AI across new sectors.