Skip to main navigation Skip to search Skip to main content

Extending OpenMP and OpenSHMEM for Efficient Heterogeneous Computing

  • Wenbin Lu
  • , Shilei Tian
  • , Tony Curtis
  • , Barbara Chapman
  • Stony Brook University

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

1 Scopus citations

Abstract

Heterogeneous supercomputing systems are becoming mainstream thanks to their powerful accelerators. However, the accelerators' special memory model and APIs increase the development complexity, and calls for innovative programming model designs. To address this issue, OpenMP has added target offloading for portable accelerator programming, and MPI allows transparent send-receive of accelerator memory buffers. Meanwhile, Partitioned Global Address Space (PGAS) languages like OpenSHMEM are falling behind for heterogeneous computing because their special memory models pose additional challenges.We propose language and runtime interoperability extensions for both OpenMP and OpenSHMEM to enable portable remote access on GPU buffers, with minimal amount of code changes. Our modified runtime systems work in coordination to manage accelerator memory, eliminating the need for staging communication buffers. Compared to the standard implementation, our extensions attain 6x point-to-point latency improvement, 1.3x better collective operation latency, 4.9x random access throughput, and up to 12.5% better performance in strong scaling configurations.

Original languageEnglish
Title of host publicationProceedings of PAW-ATM 2022
Subtitle of host publicationParallel Applications Workshop, Alternatives to MPI+X, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1-12
Number of pages12
ISBN (Electronic)9781665454100
DOIs
StatePublished - 2022
Event5th Annual IEEE/ACM Parallel Applications Workshop, Alternatives to MPI+X, PAW-ATM 2022 - Dallas, United States
Duration: Nov 13 2022Nov 18 2022

Publication series

NameProceedings of PAW-ATM 2022: Parallel Applications Workshop, Alternatives to MPI+X, Held in conjunction with SC 2022: The International Conference for High Performance Computing, Networking, Storage and Analysis

Conference

Conference5th Annual IEEE/ACM Parallel Applications Workshop, Alternatives to MPI+X, PAW-ATM 2022
Country/TerritoryUnited States
CityDallas
Period11/13/2211/18/22

Keywords

  • Heterogeneous Computing
  • Hybrid Programming
  • LLVM
  • OpenMP
  • OpenSHMEM
  • UCX

Fingerprint

Dive into the research topics of 'Extending OpenMP and OpenSHMEM for Efficient Heterogeneous Computing'. Together they form a unique fingerprint.

Cite this