Hardware Demonstration of Neural-Surrogated Quantum-Enhanced MCMC for Large-Scale Constrained Boltzmann Sampling
Abstract
Neural-surrogated quantum-enhanced Markov chain Monte Carlo (QeMCMC) uses neural networks trained on samples from the quantum approximate optimization algorithm (QAOA) to accelerate Boltzmann sampling. A divide-and-conquer strategy extends this approach to larger systems, but independently optimizing QAOA parameters for each subproblem is costly. In this paper, we develop a hardware-efficient implementation of neural-surrogated QeMCMC and demonstrate it using an IBM Quantum Heron r2 processor. Experiments with N=144 binary variables reveal faster autocorrelation decay for appropriate subproblem sizes while exposing challenges in neural-surrogate learning for larger blocks. We then extend the demonstration to a system with N=1064, where the autocorrelation decay rate per MCMC step is approximately $5.2$ and $12.3$ times those of global and local Kawasaki dynamics, respectively. These results demonstrate faster autocorrelation decay in a system exceeding one thousand variables using samples generated on quantum hardware.
Related
- Demonstration of Parallel Multi-QPU Execution for Fragment-Based Quantum Chemistry Using On-Premises Hardware
- Benchmarking One-Dimensional Quantum Lattice Boltzmann Methods on Quantum Hardware
- Experimental demonstration of response-enhanced fiber Sagnac gyroscopy via imaginary weak-value amplification
- Dephasing-Enhanced Response and Phase Transitions in Quantum Boltzmann Samplers
- Soliton Profiles in 1D: Enhanced Neural-Network Solvers and Constrained Deep Ritz Methods
BibTeX
@misc{kawamata2026hardware,
title = {{Hardware Demonstration of Neural-Surrogated Quantum-Enhanced MCMC for Large-Scale Constrained Boltzmann Sampling}},
author = {Yuya Kawamata and Keisuke Fujii},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.05883},
eprint = {2611.05883},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.05883},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Hardware Demonstration of Neural-Surrogated Quantum-Enhanced MCMC for Large-Scale Constrained Boltzmann Sampling
AU - Yuya Kawamata
AU - Keisuke Fujii
PY - 2026
DA - 2026/11/04
PB - arXiv
DO - 10.48550/arxiv.2611.05883
UR - https://arxiv.org/abs/2611.05883
AB - Neural-surrogated quantum-enhanced Markov chain Monte Carlo (QeMCMC) uses neural networks trained on samples from the quantum approximate optimization algorithm (QAOA) to accelerate Boltzmann sampling. A divide-and-conquer strategy extends this approach to larger systems, but independently optimizing QAOA parameters for each subproblem is costly. In this paper, we develop a hardware-efficient implementation of neural-surrogated QeMCMC and demonstrate it using an IBM Quantum Heron r2 processor. Experiments with N=144 binary variables reveal faster autocorrelation decay for appropriate subproblem sizes while exposing challenges in neural-surrogate learning for larger blocks. We then extend the demonstration to a system with N=1064, where the autocorrelation decay rate per MCMC step is approximately $5.2$ and $12.3$ times those of global and local Kawasaki dynamics, respectively. These results demonstrate faster autocorrelation decay in a system exceeding one thousand variables using samples generated on quantum hardware.
ER -