Exploration of Fine-Grained Parallelism for Load Balancing Eager K-truss on GPU and CPU

TL;DR

This paper introduces a fine-grained parallel approach for the Eager K-truss algorithm, significantly improving load balancing and performance on both CPU and GPU architectures.

Contribution

It presents a novel fine-grained parallel execution method for Eager K-truss, enhancing load balancing and enabling efficient GPU implementation.

Findings

Up to 1.48x speedup on CPU

Up to 16.92x speedup on GPU

Improved load balancing and parallelism

Abstract

In this work we present a performance exploration on Eager K-truss, a linear-algebraic formulation of the K-truss graph algorithm. We address performance issues related to load imbalance of parallel tasks in symmetric, triangular graphs by presenting a fine-grained parallel approach to executing the support computation. This approach also increases available parallelism, making it amenable to GPU execution. We demonstrate our fine-grained parallel approach using implementations in Kokkos and evaluate them on an Intel Skylake CPU and an Nvidia Tesla V100 GPU. Overall, we observe between a 1.261. 48x improvement on the CPU and a 9.97-16.92x improvement on the GPU due to our fine-grained parallel formulation.