Quantum Volume Benchmark

Quantum Volume Benchmark

Quantum Volume is a single-number, hardware-agnostic benchmark introduced by IBM that captures the largest random circuit of equal width and depth a quantum computer can run reliably, summarizing the combined effect of qubit count, connectivity, gate fidelity, and crosstalk into one figure of merit. This Cirq example implements the full Quantum Volume protocol end to end: it generates random model circuits of a chosen width and depth, computes their ideal output distributions and the corresponding heavy output bitstrings (those above the median probability), compiles the circuits to a target device's gate set and connectivity, runs them on a sampler (simulator or hardware), and measures the heavy-output probability. A device passes a given volume when it reproduces the heavy outputs above the two-thirds threshold with statistical confidence. The script is configurable through command-line arguments for the number of qubits, depth, number of circuits, and random seed, and returns a structured result object containing the model circuits, their heavy sets, the compiled circuits, and the sampled results, making it a practical, reproducible tool for evaluating and comparing quantum hardware.

Run it

pip install -r requirements.txt
python quantum_volume.py

Source and license

Imported from examples/advanced/quantum_volume.py in quantumlib/Cirq at v1.6.1, under the Apache License 2.0. Original authors: The Cirq Developers. The upstream LICENSE is included alongside this example.