Linear-Scaling Quantum Circuits for Computational Chemistry

TL;DR

This paper introduces optimized quantum circuits for computational chemistry that significantly reduce CNOT gate counts, maintaining accuracy and minimizing symmetry breaking, thus advancing practical quantum simulations.

Contribution

It presents approximations to existing quantum circuits that further decrease CNOT counts while preserving energy accuracy and reducing symmetry breaking.

Findings

CNOT counts are substantially reduced with approximations.

Energy accuracy remains comparable to original circuits.

Symmetry breaking is negligible with the new approximations.

Abstract

We have recently constructed compact, CNOT-efficient, quantum circuits for fermionic and qubit excitations of arbitrary many-body rank [I. Magoulas and F.A. Evangelista, J. Chem. Theory Comput. 19, 822 (2023)]. Here, we present approximations to these circuits that substantially reduce the CNOT counts even further. Our preliminary numerical data, using the selected projective quantum eigensolver approach, demonstrate that there is practically no loss of accuracy in the energies compared to the parent implementation while the ensuing symmetry breaking is essentially negligible.