Algorithm for the replica redistribution in the implementation of parallel annealing method on the hybrid supercomputer architecture

TL;DR

This paper presents an efficient implementation of the parallel annealing algorithm on hybrid CUDA and MPI architecture, demonstrating scalable performance on large-scale Ising model simulations with over two million replicas.

Contribution

It introduces a novel replica redistribution algorithm optimized for hybrid GPU-CPU architectures, enabling highly scalable parallel annealing simulations.

Findings

Acceleration approaches ideal scaling with increased system complexity

Successful simulation of over two million replicas of the Ising model

Efficient GPU utilization through optimized replica redistribution

Abstract

The parallel annealing method is one of the promising approaches for large scale simulations as potentially scalable on any parallel architecture. We present an implementation of the algorithm on the hybrid program architecture combining CUDA and MPI. The problem is to keep all general-purpose graphics processing unit devices as busy as possible redistributing replicas and to do that efficiently. We provide details of the testing on Intel Skylake/Nvidia V100 based hardware running in parallel more than two million replicas of the Ising model sample. The results are quite optimistic because the acceleration grows toward the perfect line with the growing complexity of the simulated system.