Brief Announcement: Distributed Unconstrained Local Search for Multilevel Graph Partitioning

TL;DR

This paper introduces a distributed local search algorithm for multilevel graph partitioning that significantly improves cut quality and scalability, narrowing the gap with sequential methods.

Contribution

The paper presents an engineering of an unconstrained local search algorithm for distributed graph partitioning, achieving high quality and scalability.

Findings

Scales to thousands of processing elements (PEs).

Cuts are on average 3.5% larger than state-of-the-art shared-memory partitioners.

Achieves 6.8% smaller cuts and is over 9 times faster than previous distributed methods.

Abstract

Partitioning a graph into blocks of roughly equal weight while cutting only few edges is a fundamental problem in computer science with numerous practical applications. While shared-memory parallel partitioners have recently matured to achieve the same quality as widely used sequential partitioners, there is still a pronounced quality gap between distributed partitioners and their sequential counterparts. In this work, we shrink this gap considerably by describing the engineering of an unconstrained local search algorithm suitable for distributed partitioners. We integrate the proposed algorithm in a distributed multilevel partitioner. Our extensive experiments show that the resulting algorithm scales to thousands of PEs while computing cuts that are, on average, only 3.5% larger than those of a state-of-the-art high-quality shared-memory partitioner. Compared to previous distributed partitioners, we obtain on average 6.8% smaller cuts than the best-performing competitor while being more than 9 times faster.