Source-linked AI summary
Robust randomized benchmarking of quantum processes
Easwar Magesan, J. M. Gambetta, Joseph Emerson
TL;DR
Quantum process tomography is costly and relies on unrealistically accurate preparation and measurement, motivating scalable alternatives. This paper develops randomized benchmarking models for estimating average error rates under time- and gate-dependent noise, proves validity conditions, and illustrates the models numerically.
Problem
Quantum process tomography requires exponentially many experiments and assumes preparation and measurement errors are much smaller than process errors, limiting scalable characterization.
Method
The paper applies random Clifford-operation sequences, measures average sequence fidelity, and derives perturbative zero’th- and first-order fitting models for realistic time- and gate-dependent noise.
Results
The protocol provides an efficient and reliable average error-rate estimate when error variation is not too strong, with numerical examples supporting both fitting models.
Takeaways & Limitations
Randomized benchmarking can estimate average error rates while accounting for preparation, measurement, time-dependent, and gate-dependent errors within the stated validity conditions.
Takeaways & Limitations
The perturbative analysis requires small error variation, specifically γ ≪ 2/m for neglecting second-order terms, and m ≫ 1 for sufficient fitting data.
Abstract
from arXiv · showhide
We describe a simple randomized benchmarking protocol for quantum information processors and obtain a sequence of models for the observable fidelity decay as a function of a perturbative expansion of the errors. We are able to prove that the protocol provides an efficient and reliable estimate of an average error-rate for a set operations (gates) under a general noise model that allows for both time and gate-dependent errors. We determine the conditions under which this estimate remains valid and illustrate the protocol through numerical examples.