Advertisement

The European Physical Journal Special Topics

, Volume 210, Issue 1, pp 17–31 | Cite as

An introduction to multi-GPU programming for physicists

  • M. BernaschiEmail author
  • M. Bisson
  • M. Fatica
  • E. Phillips
Review

Abstract

We present and compare different approaches for using multiple Graphics Processing Units in the simulation of physical systems. As benchmarks we consider the time required to update a single spin of the 3D Heisenberg spin glass model, by using both the Over-relaxation and the Heat Bath algorithms, and the solution of a Poisson equation by using a finite-difference method. The results show that a suitable combination of techniques allows to hide almost completely the communication overhead by using the CPU as a communication coprocessor of the GPU. Large scale simulations on clusters of GPUs can be efficiently carried out by following the same approach for other applications where a clear cut exists between bulk and boundaries data.

Keywords

Graphic Processing Unit European Physical Journal Special Topic Single Precision Graphic Processing Unit Memory Multiple GPUs 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    M.Bernaschi, G.Parisi, L.Parisi, Comput. Phys. Comm. 182, 6 (2011)CrossRefGoogle Scholar
  2. 2.
    S.Adler, Phys. Rev. D 23, 2901 (1981)ADSCrossRefGoogle Scholar
  3. 3.
    T.Preis, P.Virnau, W.Paul, J.Schneider, J. Comput. Phys. 228, 4468 (2009)ADSzbMATHCrossRefGoogle Scholar
  4. 4.
    M.Weigel, Comput. Phys. Commun. 182, 1833 (2011)ADSCrossRefGoogle Scholar
  5. 5.
    NVIDIA CUDA Compute Unified Device Architecture Programming Guide http://www.nvidia.com/cuda
  6. 6.
    I.Campos, M.Cotallo-Aban, V.Martin-Mayor, S.Perez-Gavir, A.Tarancon, Phys. Rev. Lett. 97, 217204 (2006)ADSCrossRefGoogle Scholar
  7. 7.
    The Riken Himeno CFD Benchmark: http://accc.riken.jp/HPC/HimenoBMT/index_e.html
  8. 8.
  9. 9.

Copyright information

© EDP Sciences and Springer 2012

Authors and Affiliations

  • M. Bernaschi
    • 1
    Email author
  • M. Bisson
    • 1
  • M. Fatica
    • 2
  • E. Phillips
    • 2
  1. 1.Istituto per le Applicazioni del CalcoloNational Research Council of ItalyRomeItaly
  2. 2.NVIDIA CorporationSanta ClaraUSA

Personalised recommendations