Skip to main navigation Skip to search Skip to main content

A Virtual GPU as Developer-Friendly OpenMP Offload Target

  • Atmn Patel
  • , Shilei Tian
  • , Johannes Doerfert
  • , Barbara Chapman
  • University of Waterloo
  • Stony Brook University
  • Argonne National Laboratory

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

14 Scopus citations

Abstract

While parallel programming is hard, programming accelerators has always been even more complicated. One fundamental reason is the lack of mature tooling that can be used to inspect a program that executes on two different architectures. As GPU software stacks of different vendors provide vastly different experience for developers, it is clear that the gold standard for debugging is still host (CPU) execution with its myriad of mature tooling options. In this work we present a virtual GPU (VGPU) OpenMP offloading target that allows to emulate a GPU execution environment on the host. In contrast to classical "host offloading", the VGPU target reuses the same execution model, compilation paths, and runtimes as a physical GPU. While this execution mode is not able to perform as good as host-specific compilation, runtimes, and execution, it provides the developor with a more accurate stand-in for GPU offloading that is still amendable to existing host tooling.

Original languageEnglish
Title of host publication50th International Conference on Parallel Processing Workshop, ICPP 2021 - Proceedings
PublisherAssociation for Computing Machinery
ISBN (Electronic)9781450384414
DOIs
StatePublished - Aug 9 2021
Event50th International Conference on Parallel Processing Workshop, ICPP 2021 - Virtual, Online, United States
Duration: Aug 9 2021Aug 12 2021

Publication series

NameACM International Conference Proceeding Series

Conference

Conference50th International Conference on Parallel Processing Workshop, ICPP 2021
Country/TerritoryUnited States
CityVirtual, Online
Period08/9/2108/12/21

Keywords

  • Accelerator offloading
  • Debugging
  • Gpu
  • Llvm
  • Openmp

Fingerprint

Dive into the research topics of 'A Virtual GPU as Developer-Friendly OpenMP Offload Target'. Together they form a unique fingerprint.

Cite this