Skip to content

qiboteam/qibolab

Folders and files

NameName
Last commit message
Last commit date
Sep 13, 2024
Sep 16, 2024
Mar 20, 2024
Sep 28, 2024
Oct 28, 2024
Sep 19, 2024
Aug 16, 2024
Dec 20, 2023
Oct 28, 2024
May 6, 2021
Sep 20, 2024
Aug 30, 2024
Aug 30, 2024
Sep 20, 2024
Sep 20, 2024
Nov 13, 2023

Repository files navigation

Qibolab

codecov PyPI - Version PyPI - Python Version

Qibolab is the dedicated Qibo backend for the automatic deployment of quantum circuits on quantum hardware.

Some of the key features of Qibolab are:

  • Deploy Qibo models on quantum hardware easily.
  • Create experimental drivers for custom lab setup.
  • Support multiple heterogeneous platforms.
  • Use calibration procedures from Qibocal.

Documentation

docs

The qibolab backend documentation is available at https://qibo.science/qibolab/stable/.

Minimum working example

A simple example on how to connect to a platform and use it execute a pulse sequence:

from qibolab import create_platform

# Define platform and load specific runcard
platform = create_platform("my_platform")

# Create a pulse sequence based on native gates of qubit 0
natives = platform.natives.single_qubit[0]
sequence = natives.RX() | natives.MZ()

# Connects to lab instruments using the details specified in the calibration settings.
platform.connect()

# Execute a pulse sequence
results = platform.execute([sequence], nshots=1000)

# Grab the acquired shots corresponding to
# the measurement using its pulse id.
# The ``PulseSequence`` structure is list[tuple[ChannelId, Pulse]]
# thererefore we need to index it appropriately
# to get the acquisition pulse
readout_id = sequence.acquisitions[0][1].id
print(results[readout_id])

# Disconnect from the instruments
platform.disconnect()

Arbitrary pulse sequences can also be created using the pulse API:

from qibolab import (
    Acquisition,
    Delay,
    Gaussian,
    Pulse,
    PulseSequence,
    Readout,
    Rectangular,
)

# Crete some pulses
pulse = Pulse(
    amplitude=0.3,
    duration=40,
    relative_phase=0,
    envelope=Gaussian(rel_sigma=0.2),  # Gaussian shape with std = 0.2 * duration
)
delay = Delay(duration=40)
readout = Readout(
    acquisition=Acquisition(duration=2000),
    probe=Pulse(
        amplitude=0.9,
        duration=2000,
        envelope=Rectangular(),
        relative_phase=0,
    ),
)

# Add them to a PulseSequence
sequence = PulseSequence(
    [
        (1, pulse),  # pulse plays on channel 1
        (2, delay),  # delay and readout plays on channel 2
        (2, readout),
    ]
)

Here is another example on how to execute circuits:

from qibo import gates, models, set_backend

# Create circuit and add native gates
c = models.Circuit(1)
c.add(gates.GPI2(0, phi=0.2))
c.add(gates.M(0))


# Simulate the circuit using numpy
set_backend("numpy")
result = c(nshots=1024)
print(result.probabilities())

# Execute the circuit on hardware
set_backend("qibolab", platform="my_platform")
result = c(nshots=1024)
print(result.probabilities())

Citation policy

arXiv DOI

If you use the package please refer to the documentation for citation instructions.