Parallel Local Search: Experiments with a PGAS-based programming model

TL;DR

This paper explores parallelizing local search algorithms using a PGAS-based programming model, demonstrating promising speedups and insights into their parallel behavior on modern multi-core and many-core systems.

Contribution

It introduces a new parallel version of Adaptive Search based on GPI, advancing understanding of local search parallelization on scalable architectures.

Findings

Achieved significant speedups in local search algorithms.

Provided deeper insights into parallelization of local search methods.

Validated the approach on different combinatorial problems.

Abstract

Local search is a successful approach for solving combinatorial optimization and constraint satisfaction problems. With the progressing move toward multi and many-core systems, GPUs and the quest for Exascale systems, parallelism has become mainstream as the number of cores continues to increase. New programming models are required and need to be better understood as well as data structures and algorithms. Such is the case for local search algorithms when run on hundreds or thousands of processing units. In this paper, we discuss some experiments we have been doing with Adaptive Search and present a new parallel version of it based on GPI, a recent API and programming model for the development of scalable parallel applications. Our experiments on different problems show interesting speedups and, more importantly, a deeper interpretation of the parallelization of Local Search methods.