A Comprehensive Performance Evaluation of OpenSHMEM Libraries on InfiniBand Clusters

  • Jithin Jose
  • Jie Zhang
  • Akshay Venkatesh
  • Sreeram Potluri
  • Dhabaleswar K. (DK) Panda
Part of the Lecture Notes in Computer Science book series (LNCS, volume 8356)

Abstract

OpenSHMEM is an open standard that brings together several long-standing, vendor-specific SHMEM implementations that allows applications to use SHMEM in a platform-independent fashion. Several implementations of OpenSHMEM have become available on clusters interconnected by InfiniBand networks, which has gradually become the de facto high performance network interconnect standard. In this paper, we present a detailed comparison and analysis of the performance of different OpenSHMEM implementations, using micro-benchmarks and application kernels. This study, done on TACC Stampede system using up to 4,096 cores, provides a useful guide for application developers to understand and contrast various implementations and to select the one that works best for their applications.

Keywords

OpenSHMEM Clusters InfiniBand Performance Evaluation 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Co-Array Fortran, http://www.co-array.org
  2. 2.
    Brightwell, R., Pedretti, K.: An Intra-Node Implementation of OpenSHMEM Using Virtual Address Space Mapping. In: The 5th Conference on Partitioned Global Address Space (PGAS) (2011)Google Scholar
  3. 3.
    Bonachea, D.: GASNet Specification v1.1. Tech. Rep. UCB/CSD-02-1207, U. C. Berkeley (2008)Google Scholar
  4. 4.
  5. 5.
    Jose, J., Kandalla, K., Luo, M., Panda, D.: Supporting Hybrid MPI and OpenSHMEM over InfiniBand: Design and Performance Evaluation. In: 41st International Conference on Parallel Processing, ICPP (2012)Google Scholar
  6. 6.
  7. 7.
    MVAPICH2-X: Unified MPI+PGAS Communication Runtime over OpenFabrics/Gen2 for Exascale Systems, http://mvapich.cse.ohio-state.edu/
  8. 8.
    OpenMPI: Open Source High Performance Computing, http://www.open-mpi.org/
  9. 9.
  10. 10.
  11. 11.
    Potluri, S., Kandalla, K., Bureddy, D., Li, M., Panda, D.K.: Efficient Intranode Desgins for OpenSHMEM on Multicore Clusters. In: The 6th Conference on Partitioned Global Address Space, PGAS (2012)Google Scholar
  12. 12.
    Shainer, G., Wilde, T., Lui, P., Liu, T., Kagan, M., Dubman, M., Shahar, Y., Graham, R., Shamis, P., Poole, S.: The Co-design Architecture for Exascale Systems, a Novel Approach for Scalable Designs. Computer Science-Research and Development, 1–7 (2013)Google Scholar
  13. 13.
    Silicon Graphics International: SHMEM API for Parallel Programming, http://www.shmem.org/
  14. 14.
  15. 15.
    UPC Consortium: UPC Language Specifications, v1.2. Tech. Rep. LBNL-59208, Lawrence Berkeley National Lab (2005)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2014

Authors and Affiliations

  • Jithin Jose
    • 1
  • Jie Zhang
    • 1
  • Akshay Venkatesh
    • 1
  • Sreeram Potluri
    • 1
  • Dhabaleswar K. (DK) Panda
    • 1
  1. 1.Department of Computer Science and EngineeringThe Ohio State UniversityUSA

Personalised recommendations