Scalable evaluation of incoherent infidelity in quantum devices

TL;DR

This paper introduces a scalable method to measure incoherent errors in quantum processors, providing a practical error metric that is essential for evaluating and improving quantum algorithms with potential real-world advantages.

Contribution

It presents the incoherent infidelity as a new error measure and develops a scalable technique to accurately estimate it in generic quantum evolutions under Markovian noise.

Findings

The method effectively quantifies incoherent errors in quantum circuits.

It is applicable to circuits with low error rates regardless of size.

The approach distinguishes between coherent and incoherent errors for better diagnostics.

Abstract

Quantum processors can already execute tasks beyond the reach of classical simulation, albeit for artificial problems. At this point, it is essential to design error metrics that test the experimental accuracy of quantum algorithms with potential for a practical quantum advantage. The distinction between coherent errors and incoherent errors is crucial, as they often involve different error suppression tools. The first class encompasses miscalibrations of control signals and crosstalk, while the latter is usually related to stochastic events and unwanted interactions with the environment. We introduce the incoherent infidelity as a measure of incoherent errors and present a scalable method for measuring it. This method is applicable to generic quantum evolutions subjected to time-dependent Markovian noise. Moreover, it provides an error quantifier for the target circuit, rather than an error averaged over many circuits or quantum gates. The estimation of the incoherent infidelity is suitable to assess circuits with sufficiently low error rates, regardless of the circuit size, which is a natural requirement to run useful computations.