Understanding and bridging the gaps in current GNN performance optimizations

Kezhao Huang, Jidong Zhai, Zhen Zheng, Youngmin Yi, Xipeng Shen

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

43 Scopus citations

Abstract

Graph Neural Network (GNN) has recently drawn a rapid increase of interest in many domains for its effectiveness in learning over graphs. Maximizing its performance is essential for many tasks, but remains preliminarily understood. In this work, we provide an in-depth examination of the state-of-the-art GNN frameworks, revealing five major gaps in the current frameworks in optimizing GNN performance, especially in handling the special complexities of GNN over traditional graph or DNN operations. Based on the insights, we put together a set of optimizations to fill the gaps. These optimizations leverage the state-of-the-art GPU optimization techniques and tailor them to the special properties of GNN. Experimental results show that these optimizations achieve 1.37× - 15.5× performance improvement over the state-of-the-art frameworks on various GNN models.

Original languageEnglish
Title of host publicationPPoPP 2021 - Proceedings of the 2021 26th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming
PublisherAssociation for Computing Machinery
Pages119-132
Number of pages14
ISBN (Electronic)9781450382946
DOIs
StatePublished - 17 Feb 2021
Event26th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPoPP 2021 - Virtual, Online, Korea, Republic of
Duration: 27 Feb 20213 Mar 2021

Publication series

NameProceedings of the ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPOPP

Conference

Conference26th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPoPP 2021
Country/TerritoryKorea, Republic of
CityVirtual, Online
Period27/02/213/03/21

Keywords

  • GNN
  • parallelism
  • performance optimizations

Fingerprint

Dive into the research topics of 'Understanding and bridging the gaps in current GNN performance optimizations'. Together they form a unique fingerprint.

Cite this