Customizing Graph500 for Tianhe Pre-exacale system

TL;DR

This paper enhances the Graph500 benchmark performance on the Tianhe Pre-exascale system by applying specific optimizations, achieving significant speedups and surpassing previous records with fewer nodes.

Contribution

It introduces tailored optimization techniques for distributed BFS on Tianhe Pre-exascale, including vector acceleration, buffering, and group-based communication, improving efficiency and performance.

Findings

Achieved 2131.98 GTEPS on 512 nodes, surpassing Tianhe-2 performance.

Customized Graph500 is 3.15 times faster than the base version.

Performance exceeds the June 2018 Graph500 ranking with fewer nodes.

Abstract

BFS (Breadth-First Search) is a typical graph algorithm used as a key component of many graph applications. However, current distributed parallel BFS implementations suffer from irregular data communication with large volumes of transfers across nodes, leading to inefficiency in performance. In this paper, we present a set of optimization techniques to improve the Graph500 performance for Pre-exacale system, including BFS accelerating with SVE (Scalable Vector extension) in matrix2000+, sorting with buffering for heavy vertices, and group-based monitor communication based on proprietary interconnection built in Tianhe Pre-exacale system. Performance evaluation on the customized Graph500 testing on the Tianhe Pre-exacale system achieves 2131.98 Giga TEPS on 512-node with 96608 cores, which surpasses the ranking of Tianhe-2 with about 16X fewer nodes in the June 2018 Graph500 list, and shows our customized Graph500 is 3.15 times faster on 512 nodes than the base version using the state-of-the-art techniques.