Energy Efficient Runtime Framework for Exascale Systems

Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9945)

Abstract

Building an Exascale computer that solves scientific problems by three orders of magnitude faster as the current Petascale systems is harder than just making it huge. Towards the first Exascale computer, energy consumption has been emerged to a crucial factor. Every component will have to change to create an Exascale syestem, which capable of a million trillion of computing per second. To run efficiently on these huge systems and to take advantages of every computational power, software and underlying algorithms should be rewritten. While many computing intensive applications are designed to use Message Passing Interface (MPI) with two-sided communication semantics, a Partitioned Global Address Space (PGAS) is being designed, through providing an abstraction of the global address space, to treat a distributed system as if the memory were shared. The data locality and communication could be optimized through the one sided communication offered by PGAS. In this paper we present an energy aware runtime framework, which is PGAS based and offers MPI as a substrate communication layer.

Keywords

Exascale Energy efficiency Data locality PGAS Runtime system MPI 

References

  1. 1.
  2. 2.
  3. 3.
    OSU MPI Benchmarks: OSU MVPICH. http://mvapich.cse.ohio-state.edu/benchmarks/
  4. 4.
    BwUniCluster: BwUniCluster. http://www.bwhpc-c5.de/wiki/index.php
  5. 5.
    Chetsa, G.L.T., Lefevre, L., Pierson, J.M., Stolf, P., Da Costa, G.: A runtime framework for energy efficient HPC systems without a priori knowledge of applications. In: Proceedings of the International Conference on Parallel and Distributed Systems - ICPADS, pp. 660–667 (2012)Google Scholar
  6. 6.
    Daily, J., Vishnu, A., Palmer, B., Van Dam, H., Kerbyson, D.: On the suitability of MPI as a PGAS runtime. In: 2014 21st International Conference on High Performance Computing, HiPC 2014 (2015)Google Scholar
  7. 7.
    Dinan, J., Balaji, P., Hammond, J.R., Krishnamoorthy, S., Tipparaju, V.: Supporting the global arrays PGAS model using MPI one-sided communication. In: Proceedings of the 2012 IEEE 26th International Parallel and Distributed Processing Symposium, IPDPS 2012, pp. 739–750 (2012)Google Scholar
  8. 8.
    El-Ghazawi, T., Carlson, W., Sterling, T., Yelick, K.: UPC: Distributed Shared Memory Programming. Wiley, New York (2005)CrossRefGoogle Scholar
  9. 9.
  10. 10.
    Fürlinger, K., Glass, C., Gracia, J., Knüpfer, A., Tao, J., Hünich, D., Idrees, K., Maiterth, M., Mhedheb, Y., Zhou, H.: DASH: data structures and algorithms with support for hierarchical locality. In: Lopes, L., et al. (eds.) Euro-Par 2014. LNCS, vol. 8805, pp. 542–552. Springer, Heidelberg (2014). doi:10.1007/978-3-319-14313-2_46 Google Scholar
  11. 11.
  12. 12.
    Kandalla, K., Mancini, E.P., Sur, S., Panda, D.K.: Designing power-aware collective communication algorithms for infiniband clusters. In: Proceedings of the International Conference on Parallel Processing, pp. 218–227 (2010)Google Scholar
  13. 13.
    Krawezik, G., Cappello, F.: Performance comparison of MPI and OpenMP on shared memory multiprocessors. Concurr. Comput. Pract. Exp. 18(1), 29–61 (2006)CrossRefGoogle Scholar
  14. 14.
    Mametjanov, A., Min, M., Norris, B., Hovland, P.D.: Accelerating performance of NekCEM with MPI and CUDA. In: Super Computing (SC13) (2013)Google Scholar
  15. 15.
    Mhedheb, Y., Streit, A.: Energy-efficient Task Scheduling in Data Centers (2014)Google Scholar
  16. 16.
  17. 17.
    Shmem: OpenSHMEM. http://openshmem.org/site/
  18. 18.
    SPPEXA: Software for Exascale SPPEXA. http://www.sppexa.de
  19. 19.
  20. 20.
    Vishnu, A., Song, S., Marquez, A., Barker, K., Kerbyson, D., Cameron, K., Balaji, P.: Designing energy efficient communication runtime systems: a view from PGAS models. J. Supercomput. 63(3), 691–709 (2013)CrossRefGoogle Scholar
  21. 21.
    Yang, C., Bland, W., Mellor-Crummey, J., Balaji, P.: Portable, MPI-interoperable coarray fortran. In: Proceedings of the 19th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming - PPopp 2014, pp. 81–92 (2014)Google Scholar
  22. 22.
    Zhou, H., Idrees, K., Gracia, J.: Leveraging MPI-3 shared-memory extensions for efficient PGAS runtime systems. In: Träff, J.L., Hunold, S., Versaci, F. (eds.) Euro-Par 2015. LNCS, vol. 9233, pp. 373–384. Springer, Heidelberg (2015). doi:10.1007/978-3-662-48096-0_29 CrossRefGoogle Scholar
  23. 23.
    Zhou, H., Mhedheb, Y., Idrees, K., Glass, C.W.: DART-MPI: an MPI-based implementation of a PGAS runtime system. In: PGAS 2014 (2014)Google Scholar

Copyright information

© Springer International Publishing AG 2016

Authors and Affiliations

  1. 1.Steinbuch Centre for ComputingKarlsruhe Institute of TechnologyKarlsruheGermany

Personalised recommendations