An Evaluation of Implementation Options for MPI One-Sided Communication

  • William Gropp
  • Rajeev Thakur
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3666)

Abstract

MPI defines one-sided communication operations—put, get, and accumulate—together with three different synchronization mechanisms that define the semantics associated with the initiation and completion of these operations. In this paper, we analyze the requirements imposed by the MPI Standard on any implementation of one-sided communication. We discuss options for implementing the synchronization mechanisms and analyze the cost associated with each. An MPI implementer can use this information to select the implementation method that is best suited (has the lowest cost) for a particular machine environment. We also report on experiments we ran on a Linux cluster and a Sun SMP to determine the gap between the performance that could be achievable and what is actually achieved with MPI.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Asai, N., Kentemich, T., Lagier, P.: MPI-2 implementation on Fujitsu generic message passing kernel. In: Proc. of SC 1999: High Performance Networking and Computing (November 1999)Google Scholar
  2. 2.
    Booth, S., Mourão, E.: Single sided MPI implementations for SUN MPI. In: Proc. of SC 2000: High Performance Networking and Computing (November 2000)Google Scholar
  3. 3.
    Gabriel, E., Fagg, G.E., Dongarra, J.J.: Evaluating dynamic communicators and one-sided operations for current MPI libraries. Int’l Journal of High-Performance Computing Applications 19(1), 67–80 (2005)CrossRefGoogle Scholar
  4. 4.
    Golebiewski, M., Träff, J.L.: MPI-2 One-Sided Communications on a Giganet SMP Cluster. In: Proc. of the 8th European PVM/MPI Users’ Group Meeting, September 2001, pp. 16–23 (2001)Google Scholar
  5. 5.
    Hensgen, D., Finkel, R., Manbet, U.: Two algorithms for barrier synchronization. International Journal of Parallel Programming 17(1), 1–17 (1988)MATHCrossRefGoogle Scholar
  6. 6.
    Jiang, W., Liu, J., Jin, H.-W., Panda, D.K., Buntinas, D., Thakur, R., Gropp, W.: Efficient implementation of MPI-2 passive one-sided communication over InfiniBand clusters. In: Proc. of the 11th European PVM/MPI Users’ Group Meeting, September 2004, pp. 68–76 (2004)Google Scholar
  7. 7.
    Jiang, W., Liu, J., Jin, H.-W., Panda, D.K., Gropp, W., Thakur, R.: High performance MPI-2 one-sided communication over InfiniBand. In: Proc. of the 4th Int’l Symp. on Cluster Computing and the Grid (CCGrid 2004) (April 2004)Google Scholar
  8. 8.
    Message Passing Interface Forum. MPI-2: Extensions to the Message-Passing Interface (July 1997), http://www.mpi-forum.org/docs/docs.html
  9. 9.
    Mourão, E., Booth, S.: Single sided communications in multi-protocol MPI. In: Proc. of the 7th European PVM/MPI Users’ Group Meeting, September 2000, pp. 176–183 (2000)Google Scholar
  10. 10.
    Mourão, F.E., Silva, J.G.: Implementing MPI’s one-sided communications for WMPI. In: Proc. of the 6th European PVM/MPI Users’ Group Meeting, September 1999, pp. 231–238 (1999)Google Scholar
  11. 11.
  12. 12.
    Thakur, R., Gropp, W., Toonen, B.: Optimizing the Synchronization Operations in MPI One-Sided Communication. Int’l Journal of High-Performance Computing Applications 19(2), 119–128 (2005)CrossRefGoogle Scholar
  13. 13.
    Thakur, R., Rabenseifner, R., Gropp, W.: Optimization of collective communication operations in MPICH. Int’l Journal of High-Performance Computing Applications 19(1), 49–66 (2005)CrossRefGoogle Scholar
  14. 14.
    Träff, J.L., Ritzdorf, H., Hempel, R.: The implementation of MPI-2 one-sided communication for the NEC SX-5. In: Proc. of SC 2000: High Performance Networking and Computing (November 2000)Google Scholar
  15. 15.
    Worringen, J., Gäer, A., Reker, F.: Exploiting transparent remote memory access for non-contiguous and one-sided-communication. In: Proc. of the 2002 Workshop on Communication Architecture for Clusters (April 2002)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2005

Authors and Affiliations

  • William Gropp
    • 1
  • Rajeev Thakur
    • 1
  1. 1.Mathematics and Computer Science DivisionArgonne National LaboratoryArgonneUSA

Personalised recommendations