English

Quantifying the Impact of Precision Errors on Quantum Approximate Optimization Algorithms

Quantum Physics 2021-11-11 v2

Abstract

The quantum approximate optimization algorithm (QAOA) is a hybrid quantum-classical algorithm that seeks to achieve approximate solutions to optimization problems by iteratively alternating between intervals of controlled quantum evolution. Here, we examine the effect of analog precision errors on QAOA performance both from the perspective of algorithmic training and canonical state- and observable-dependent QAOA-relevant metrics. Leveraging cumulant expansions, we recast the faulty QAOA as a control problem in which precision errors are expressed as multiplicative control noise and derive bounds on the performance of QAOA. We show using both analytical techniques and numerical simulations that errors in the analog implementation of QAOA circuits hinder its performance as an optimization algorithm. In particular, we find that any fixed precision implementation of QAOA will be subject to an exponential degradation in performance dependent upon the number of optimal QAOA layers and magnitude of the precision error. Despite this significant reduction, we show that it is possible to mitigate precision errors in QAOA via digitization of the variational parameters, therefore at the cost of increasing circuit depth. We illustrate our results via numerical simulations and analytic and empirical error bounds as a comparison. While focused on precision errors, our approach naturally lends itself to more general noise scenarios and the calculation of error bounds on QAOA performance and broader classes of variational quantum algorithms.

Keywords

Cite

@article{arxiv.2109.04482,
  title  = {Quantifying the Impact of Precision Errors on Quantum Approximate Optimization Algorithms},
  author = {Gregory Quiroz and Paraj Titum and Phillip Lotshaw and Pavel Lougovski and Kevin Schultz and Eugene Dumitrescu and Itay Hen},
  journal= {arXiv preprint arXiv:2109.04482},
  year   = {2021}
}

Comments

21 pages, 11 figures. Newest version contains updated figures and revisions to the main text and appendices

R2 v1 2026-06-24T05:50:18.873Z