Fast BFS-Based Triangle Counting on GPUs

TL;DR

This paper introduces a GPU-optimized BFS-based triangle counting method that significantly outperforms existing CPU and GPU approaches, demonstrating high scalability and efficiency for large graphs.

Contribution

A novel BFS-based GPU algorithm for triangle counting that achieves superior scalability and performance compared to prior methods.

Findings

Achieves nearly 10 GTEPS in runtime.

Outperforms all existing GPU and CPU implementations.

Demonstrates a 3.84x speedup on the Subgraph Isomorphism Graph Challenge 2019.

Abstract

In this paper, we propose a novel method to compute triangle counting on GPUs. Unlike previous formulations of graph matching, our approach is BFS-based by traversing the graph in an all-source-BFS manner and thus can be mapped onto GPUs in a massively parallel fashion. Our implementation uses the Gunrock programming model and we evaluate our implementation in runtime and memory consumption compared with previous state-of-the-art work. We sustain a peak traversed-edges-per-second (TEPS) rate of nearly 10 GTEPS. Our algorithm is the most scalable and parallel among all existing GPU implementations and also outperforms all existing CPU distributed implementations. This work specifically focuses on leveraging our implementation on the triangle counting problem for the Subgraph Isomorphism Graph Challenge 2019, demonstrating a geometric mean speedup over the 2018 champion of 3.84x.