Quantum-Enhanced Simulated Annealing Using Rydberg Atoms

TL;DR

This paper demonstrates that a Rydberg atom-based quantum-enhanced simulated annealing algorithm can outperform classical simulated annealing in solving the maximum independent set problem, showing potential for quantum advantage in complex optimization tasks.

Contribution

The study introduces a novel quantum-classical hybrid algorithm using Rydberg atoms, experimentally showing its computational time advantage over classical methods for combinatorial optimization.

Findings

QESA outperforms classical SA in approximation ratio and Hamming distance.

Experimental data from Rydberg atomic arrays validate the quantum advantage.

Estimated maximum graph size manageable within one day on a personal computer.

Abstract

Quantum-classical hybrid algorithms offer a promising strategy for tackling computationally challenging problems, such as the maximum independent set (MIS) problem that plays a crucial role in areas like network design and data analysis. This study experimentally demonstrates that a Rydberg quantum-classical hybrid algorithm, termed as quantum-enhanced simulated annealing (QESA), provides a computational time advantage over standalone simulated annealing (SA), a classical heuristic optimization method. The performance of QESA is evaluated based on the approximation ratio and the Hamming distance, relative to the graph size. The analysis shows that QESA outperforms standalone SA by leveraging a warm-start input derived from two types of Rydberg atomic array experimental data: quench evolution (QE) (implemented on the Quera Aquila machine) and adiabatic quantum computing (AQC) (using the experimental dataset archieved in K. Kim et al., Scientific Data 11, 111 (2024). Based on these results, an estimate is provided for the maximum graph size that can be handled within a one-day computational time limit on a standard personal computer. These findings suggest that QESA has the potential to offer a computational advantage over classical methods for solving complex optimization problems efficiently.