Define a beltrami CPU baseline

[johnmauff]

This PR adds a CPU-based Beltrami baseline to verify against future code modifications.

fix #56

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 32.62%. Comparing base (6e2fd8e) to head (a39bfcf).
Report is 14 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main      #70   +/-   ##
=======================================
  Coverage   32.62%   32.62%           
=======================================
  Files          51       51           
  Lines       16815    16815           
=======================================
  Hits         5486     5486           
  Misses      11329    11329           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[johnmauff]

I just realized an issue with this baseline. It was collected on Derecho using 128 threads. When compared with the CI/CD result, it will be different because it ran on a different number of threads resulting in non B4B results.

[sjsprecious]

I just realized an issue with this baseline. It was collected on Derecho using 128 threads. When compared with the CI/CD result, it will be different because it ran on a different number of threads resulting in non B4B results.

Thanks @johnmauff for pointing out the difference of thread numbers. I think we won't match the numbers B4B anyway because we use Intel compiler on Derecho while the CI workflow uses the GCC compiler. Probably we should target at at certain tolerance level instead?

[cenamiller]

Can we rename this to a more accessible generic name? Thank you

[cenamiller]

Actually close this PR and issue a new one to the thermo_asap branch please

[johnmauff]

This PR will be closed as it applies to the main branch. Instead, PR #95 is its replacement.