Fast Estimation of Sparse Quantum Noise

As quantum computers approach the fault-tolerance threshold, diagnosing and characterizing noise on large-scale devices is increasingly important. One of most important classes channels class Pauli channels, for reasons both theoretical tractability experimental relevance. Here we present a practical algorithm estimating s nonzero error rates in an s-sparse, n-qubit channel, or more generally largest rates. The comes with rigorous recovery guarantees uses only O(n2) measurements, O(sn2) classical processing time, Clifford circuits. We experimentally validate heuristic version that simplified circuits data from IBM 14-qubit superconducting device our open-source implementation. These show accurate precise estimation probability arbitrary-weight errors possible even when signal 2 orders magnitude below measurement floor.Received August 2020Revised 22 December 2020Accepted 13 January 2021DOI:https://doi.org/10.1103/PRXQuantum.2.010322Published by American Physical Society under terms Creative Commons Attribution 4.0 International license. Further distribution this work must maintain attribution to author(s) published article's title, journal citation, DOI.Published SocietyPhysics Subject Headings (PhySH)Research AreasQuantum benchmarkingQuantum channelsQuantum tomographyQuantum Information