A blueprint for a Digital-Analog Variational Quantum Eigensolver using Rydberg atom arrays

TL;DR

This paper proposes a digital-analog variational quantum eigensolver tailored for Rydberg atom arrays to efficiently estimate molecular ground-state energies, combining atom position learning and pulse optimization.

Contribution

It introduces a novel digital-analog VQE approach using Rydberg atoms, integrating atom position learning and iterative pulse shaping for improved energy estimation.

Findings

Achieved energy estimates within a few percent error for molecules like H2, LiH, BeH2.

Demonstrated the effectiveness of atom position learning in the VQE process.

Validated the approach through numerical simulations on molecular Hamiltonians.

Abstract

We address the task of estimating the ground-state energy of Hamiltonians coming from chemistry. We study numerically the behavior of a digital-analog variational quantum eigensolver for the H2, LiH and BeH2 molecules, and we observe that one can estimate the energy to a few percent points of error leveraging on learning the atom register positions with respect to selected features of the molecular Hamiltonian and then an iterative pulse shaping optimization, where each step performs a derandomization energy estimation.