Continuous monitoring for noisy intermediate-scale quantum processors

TL;DR

This paper introduces a continuous monitoring system for intermediate-scale quantum processors that estimates noise characteristics from executed circuits without additional control, aiding calibration and benchmarking.

Contribution

The paper presents a novel monitoring approach that extracts noise estimates from existing circuit data, independent of control over the input circuits, applicable to real and simulated quantum data.

Findings

Effective noise estimation from existing circuit data

Validated on both quantum emulator and real quantum processor

Potential to reduce resources for quantum device calibration

Abstract

We present a continuous monitoring system for intermediate-scale quantum processors that allows extracting estimates of noisy native gate and read-out measurements based on the set of executed quantum circuits and resulting measurement outcomes. In contrast to standard approaches for calibration and benchmarking quantum processors, the executed circuits, which are input to the monitoring system, are assumed to be out of any control. We provide the results of applying our system to the synthetically generated data obtained from a quantum emulator, as well as to the experimental data collected from a publicly accessible cloud-based quantum processor. In the both cases, we demonstrate that the developed approach provides valuable results about inherent noises of emulators/processors. Considering that our approach uses only already accessible data from implemented circuits without the need to run additional algorithms, the monitoring system can complement existing approaches. We expect that our monitoring system can become a useful tool for various quantum computers in the near-term horizon, including publicly accessible cloud-based platforms, and reduce resources that are required for their benchmarking and calibration.