Decoupling Schedule, Topology Layout, and Algorithm to Easily Enlarge the Tuning Space of GPU Graph Processing

Shinnung Jeong, Yongwoo Lee, Jaeho Lee, Heelim Choi, Seungbin Song, Jinho Lee, Youngsok Kim, Hanjun Kim

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Only with a right schedule and a right topology layout, a graph algorithm can be efciently processed on GPUs. Existing GPU graph processing frameworks try to fnd an optimal schedule and topology layout for an algorithm via iterative search, but they fail to fnd the optimal confguration because their schedules and topology layouts are tightly coupled in their processing models. Moreover, their tightly coupled schedules and topology layouts make it diffcult for developers to extend the tuning space. To easily enlarge the tuning space of GPU graph processing, this work proposes a new GPU graph processing abstraction scheme that fully decouples schedules, topology layouts, and algorithms from each other with abstraction interfaces. Moreover, this work proposes GRAssembler, a new GPU graph processing framework that efciently integrates the decoupled schedule, topology layout, and algorithm without abstraction overhead. Thanks to the efcient decoupling and integration, GRAssembler increases the tuning space from 336 to 4,480 and achieves 30.4% higher performance on geomean average, compared to the state-of-the-art GPU graph processing framework.

Original languageEnglish
Title of host publicationPACT 2022 - Proceedings of the 2022 International Conference on Parallel Architectures and Compilation Techniques
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages198-210
Number of pages13
ISBN (Electronic)9781450398688
DOIs
Publication statusPublished - 2022 Oct 8
Event31st International Conference on Parallel Architectures and Compilation Techniques, PACT 2022 - Chicago, United States
Duration: 2022 Oct 82022 Oct 10

Publication series

NameParallel Architectures and Compilation Techniques - Conference Proceedings, PACT
ISSN (Print)1089-795X

Conference

Conference31st International Conference on Parallel Architectures and Compilation Techniques, PACT 2022
Country/TerritoryUnited States
CityChicago
Period22/10/822/10/10

Bibliographical note

Publisher Copyright:
© 2022 Association for Computing Machinery.

All Science Journal Classification (ASJC) codes

  • Software
  • Theoretical Computer Science
  • Hardware and Architecture

Fingerprint

Dive into the research topics of 'Decoupling Schedule, Topology Layout, and Algorithm to Easily Enlarge the Tuning Space of GPU Graph Processing'. Together they form a unique fingerprint.

Cite this