ELF: Maximizing memory-level parallelism for GPUs with coordinated warp and fetch scheduling

Jason Jong Kyu Park, Yongjun Park, Scott Mahlke

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

9 Citations (Scopus)

Abstract

Graphics processing units (GPUs) are increasingly utilized as throughput engines in the modern computer systems. GPUs rely on fast context switching between thousands of threads to hide long latency operations, however, they still stall due to the memory operations. To minimize the stalls, memory operations should be overlapped with other operations as much as possible to maximize memory-level parallelism (MLP). In this paper, we propose Earliest Load First (ELF) warp scheduling, which maximizes the MLP by giving higher priority to the warps that have the fewest instructions to the next memory load. ELF utilizes the same warp priority for the fetch scheduling so that both are coordinated. We also show that ELF reveals its full benefits when there are fewer memory conflicts and fetch stalls. Evaluations show that ELF can improve the performance by 4.1% and achieve total improvement of 11.9% when used with other techniques over commonly-used greedy-then-oldest scheduling.

Original languageEnglish
Title of host publicationProceedings of SC 2015
Subtitle of host publicationThe International Conference for High Performance Computing, Networking, Storage and Analysis
PublisherIEEE Computer Society
ISBN (Electronic)9781450337236
DOIs
Publication statusPublished - 2015 Nov 15
EventInternational Conference for High Performance Computing, Networking, Storage and Analysis, SC 2015 - Austin, United States
Duration: 2015 Nov 152015 Nov 20

Publication series

NameInternational Conference for High Performance Computing, Networking, Storage and Analysis, SC
Volume15-20-November-2015
ISSN (Print)2167-4329
ISSN (Electronic)2167-4337

Conference

ConferenceInternational Conference for High Performance Computing, Networking, Storage and Analysis, SC 2015
Country/TerritoryUnited States
CityAustin
Period15/11/1515/11/20

Bibliographical note

Funding Information:
We would like to thank the anonymous reviewers as well as the fellow members of the CCCP research group for their valuable comments and feedbacks. This work is supported in part by the National Science Foundation under grant SHF-1217917 and by the Defense Advanced Research Projects Agency (DARPA) under the Power Efficiency Revolution for Embedded Computing Technologies (PERFECT) program.

Publisher Copyright:
© 2015 ACM.

All Science Journal Classification (ASJC) codes

  • Computer Networks and Communications
  • Computer Science Applications
  • Hardware and Architecture
  • Software

Fingerprint

Dive into the research topics of 'ELF: Maximizing memory-level parallelism for GPUs with coordinated warp and fetch scheduling'. Together they form a unique fingerprint.

Cite this