Presentation

DescriptionWe present a parallelization scheme for classical simulations of quantum circuits. Our scheme is based on a recent method to "cut'' large quantum circuits into smaller sub-circuits that can be simulated independently, and whose simulation results can in turn be re-combined to infer the output of the original circuit. The exponentially smaller classical computing resources needed to simulate smaller circuits are counterbalanced by exponential overhead in terms of classical post-processing costs. We discuss how this overhead can be massively parallelized to reduce classical computing costs.

Archive

Authors

Michael A. Perlin

University of Colorado

National Institute of Standards and Technology (NIST)

Teague Tomesh

Princeton University

Bradley Pearlman

University of Colorado

National Institute of Standards and Technology (NIST)

Wei Tang

Princeton University

Yuri Alexeev

Argonne National Laboratory

Martin Suchara

Argonne National Laboratory