Towards unification of accelerated computing and interconnection for extreme-scale computing

Toshihiro Hanawa, Yuetsu Kodama, Taisuke Boku, Hideharu Amano, Hitoshi Murai, Masayuki Umemura, Mitsuhisa Sato

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

Abstract

Heterogeneous clusters using accelerators are widely used for high-performance computing system. In such systems, the inter-node communication among accelerators becomes bottleneck due to the data transfer between the accelerator and the host. To eliminate this overhead, we have been developing a novel communication system realizing direct communication among accelerators over computation nodes under the HA-PACS (Highly Accelerated Parallel Advanced system for Computational Sciences) project. Also we are investigating high-level parallel programming language, and several practical application programs on our concept, as well as studying the enhancement of TCA and developing system software stack in the CREST project.

Original languageEnglish
Title of host publicationApplied Reconfigurable Computing - 11th International Symposium, ARC 2015, Proceedings
EditorsPedro C. Diniz, Kentaro Sano, Michael Hübner, Dimitrios Soudris
PublisherSpringer Verlag
Pages463-474
Number of pages12
ISBN (Electronic)9783319162133
DOIs
Publication statusPublished - 2015 Jan 1
Event11th International Symposium on Applied Reconfigurable Computing, ARC 2015 - Bochum, Germany
Duration: 2015 Apr 132015 Apr 17

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9040
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other11th International Symposium on Applied Reconfigurable Computing, ARC 2015
CountryGermany
CityBochum
Period15/4/1315/4/17

Keywords

  • Accelerator computing
  • GPU direct
  • Interconnection network
  • PCI express
  • Remote DMA

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Towards unification of accelerated computing and interconnection for extreme-scale computing'. Together they form a unique fingerprint.

  • Cite this

    Hanawa, T., Kodama, Y., Boku, T., Amano, H., Murai, H., Umemura, M., & Sato, M. (2015). Towards unification of accelerated computing and interconnection for extreme-scale computing. In P. C. Diniz, K. Sano, M. Hübner, & D. Soudris (Eds.), Applied Reconfigurable Computing - 11th International Symposium, ARC 2015, Proceedings (pp. 463-474). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9040). Springer Verlag. https://doi.org/10.1007/978-3-319-16214-0_43