DMA-assisted, intranode communication in GPU accelerated systems

Feng Ji*, Ashwin M. Aji, James Dinan, Darius Buntinas, Pavan Balaji, Rajeev Thakur, Wu Chun Feng, Xiaosong Ma

*Corresponding author for this work

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

8 Citations (Scopus)

Abstract

Accelerator awareness has become a pressing issue in data movement models, such as MPI, because of the rapid deployment of systems that utilize accelerators. In our previous work, we developed techniques to enhance MPI with accelerator awareness, thus allowing applications to easily and efficiently communicate data between accelerator memories. In this paper, we extend this work with techniques to perform efficient data movement between accelerators within the same node using a DMA-assisted, peer-to-peer intranode communication technique that was recently introduced for NVIDIA GPUs. We present a detailed design of our new approach to intranode communication and evaluate its improvement to communication and application performance using micro-kernel benchmarks and a 2D stencil application kernel.

Original languageEnglish
Title of host publicationProceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012
Pages461-468
Number of pages8
DOIs
Publication statusPublished - 2012
Externally publishedYes
Event14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012 - Liverpool, United Kingdom
Duration: 25 Jun 201227 Jun 2012

Publication series

NameProceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012

Conference

Conference14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012
Country/TerritoryUnited Kingdom
CityLiverpool
Period25/06/1227/06/12

Keywords

  • GPU
  • Intranode communication
  • MPI

Fingerprint

Dive into the research topics of 'DMA-assisted, intranode communication in GPU accelerated systems'. Together they form a unique fingerprint.

Cite this