Characterizing quantum processors using discrete time crystals

TL;DR

This paper introduces a novel method using discrete time crystals to evaluate quantum processor performance, revealing variability across devices and layouts, and emphasizing the importance of full-device interrogation for accurate benchmarking.

Contribution

The authors develop a platform-agnostic, scalable benchmarking technique based on discrete time crystals that captures true quantum processor performance and identifies weak regions at the qubit level.

Findings

Significant variability in scores across different quantum processors.

Full-device interrogation provides more accurate performance assessment.

Infrequent benchmarks may not reflect real-world capabilities.

Abstract

We present a method for characterizing the performance of noisy quantum processors using discrete time crystals. Deviations from ideal persistent oscillatory behavior give rise to numerical scores by which relative quantum processor capabilities can be measured. We construct small sets of qubit layouts that cover the full topology of a target system, and execute our metric over these sets on a wide range of IBM Quantum processors. We show that there is a large variability in scores, not only across multiple processors, but between different circuit layouts over individual devices. The stability of results is also examined. Our results suggest that capturing the true performance characteristics of a quantum system requires interrogation over the full device, rather than isolated subgraphs. Moreover, the disagreement between our results and other metrics indicates that benchmarks computed infrequently are not indicative of the real-world performance of a quantum processor. This method is platform agnostic, simple to implement, and scalable to any number of qubits forming a linear-chain, while simultaneously allowing for identifying ill-performing regions of a device at the individual qubit level.