Accelerating Scientific Applications on Heterogeneous Systems with HybridOMP

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

Abstract

High Performance Computing relies on accelerators (such as GPGPUs) to achieve fast execution of scientific applications. Traditionally these accelerators have been programmed with specialized languages, such as CUDA or OpenCL. In recent years, OpenMP emerged as a promising alternative for supporting accelerators, providing advantages such as maintaining a single code base for the host and different accelerator types and providing a simple way to extend support for accelerators to existing code. Efficiently using this support requires solving several challenges, related to performance, work partitioning, and concurrent execution on multiple device types. In this paper, we discuss these challenges and introduce a library, HybridOMP, that addresses several of them, thus enabling the effective use of OpenMP for accelerators. We apply HybridOMP to a scientific application, PlasCom2, that has not previously been able to use accelerators. Experiments on three architectures show that HybridOMP results in performance gains of up to 10x compared to CPU-only execution. Concurrent execution on the host and GPU resulted in additional gains of up to 10% compared to running on the GPU only.

Original languageEnglish (US)
Title of host publicationHigh Performance Computing for Computational Science – VECPAR 2018 - 13th International Conference, Revised Selected Papers
EditorsHermes Senger, Osni Marques, Tatiana Pinheiro de Brito, Rogério Iope, Silvio Stanzani, Veronica Gil-Costa, Rogerio Garcia
PublisherSpringer-Verlag Berlin Heidelberg
Pages174-187
Number of pages14
ISBN (Print)9783030159955
DOIs
StatePublished - Jan 1 2019
Event13th International Conference on High Performance Computing in Computational Science, VECPAR 2018 - São Pedro, Brazil
Duration: Sep 17 2018Sep 19 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11333 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference13th International Conference on High Performance Computing in Computational Science, VECPAR 2018
CountryBrazil
CitySão Pedro
Period9/17/189/19/18

    Fingerprint

Keywords

  • Accelerators
  • GPGPU
  • Heterogeneous computing
  • OpenMP

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Diener, M., Bodony, D. J., & Kale, L. (2019). Accelerating Scientific Applications on Heterogeneous Systems with HybridOMP. In H. Senger, O. Marques, T. P. de Brito, R. Iope, S. Stanzani, V. Gil-Costa, & R. Garcia (Eds.), High Performance Computing for Computational Science – VECPAR 2018 - 13th International Conference, Revised Selected Papers (pp. 174-187). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11333 LNCS). Springer-Verlag Berlin Heidelberg. https://doi.org/10.1007/978-3-030-15996-2_13