Skip to main content
SpringerLink
Log in

On partitioning and mapping for hypercube computing

  • Published:
International Journal of Parallel Programming Aims and scope Submit manuscript

Abstract

Designing efficient parallel algorithms in a message-based parallel computer should consider both time-space tradeoffs and computation-communication tradeoffs. In order to balance these tradeoffs and achieve the optimal performance of an algorith, one has to consider various design parameters such as the number of processors required and the size of partitions. In this paper, we demonstrate that, for certain data parallel algorithms, it is possible to determine these design parameters analytically. To serve as a basis for the discussions that follow, a simple model for the NCUBE hypercube computer is introduced. Using this model, we use two examples, array summation and matrix multiplication, to illustrate how their performance can be modeled. By optimizing these expressions, one is able to determine optimal design parameters which arrive at efficient execution. Experiments on a 64-node NCUBE verified the accuracy of the analytic results and are used to further support the discussions.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Similar content being viewed by others

References

  1. W. L. Athas and C. L. Seitz, Multicomputers: Message-Passing Concurrent Computers,IEEE Computers, pp. 9–24 (August 1988).

  2. R. Olson, Parallel Processing in a Message-Based Operating System,IEEE Software, pp. 39–49 (July 1985).

  3. W. Crowther, J. Goodhue, E. Starr, R. Thomas, W. Milliken, and T. Blackadar, Performance Measurements on a 128-Node Butterfly Parallel Processor,Proc. of 1985 Int'l Conf. on Parallel Processing, pp. 531–540 (August 1985).

  4. W. D. Hillis and G. L. Steele, Data Parallel Algorithms,Comm. ACM, pp. 1170–1183 (December 1986).

  5. M. R. Garey and D. S. Johnson, Computers and Intractability: A Guide to the Theory of NP-Completeness, W. H. Freeman and Co. (1979).

  6. F. Berman, Experience with an Automatic Solution to the Mapping Problem, inThe Characteristics of Parallel Algorithms, L. H. Jamieson, D. B. Gannon, R. J. Douglass, Eds., MIT Press (1987).

  7. Y. Saad and M. H. Schultz, Topological Properties of Hypercubes, Technical Report, YALEU/DCS/RR-389, Department of Computer Science, Yale University (June 1985).

  8. W. J. Dally and C. L. Seitz, Deadlock-Free Message Routing in Multiprocessor Interconnection Networks,IEEE Trans. on Computers, pp. 547–553 (May, 1987).

  9. D. C. Grunwald and D. A. Reed, Networks for Parallel Processors: Measurements and Prognostications,Proc. of Third Conf. on Hypercube Multiprocessors, pp. 610–619 (1988).

  10. L. M. Ni, C. T. King, and P. Prins, Parallel Algorithm Design Considerations for Hypercube Multiprocessors,Proc. of Int'l Conf. on Parallel Processing, pp. 717–720 (1987).

  11. P. R. Ma, E. Y. Lee, and M. Tsuchiya, A Task Allocation Model for Distributed Computing Systems,IEEE Trans. Computers, pp. 41–47 (January 1982).

  12. G. S. Rao, H. S. Stone, and T. C. Hu, Assignment of Tasks in a Distributed Processor System with Limited Memory,IEEE Trans. Computers, pp. 291–298 (April 1979).

  13. J. P. Hayes, T. N. Mudge, Q. F. Stout, S. Colley, and J. Palmer, Architecture of a Hypercube Supercomputer,Proc. of 1986 Int'l Conf. on Parallel Processing, pp. 653–660 (August 1986).

  14. D. C. Grunwald, D. A. Reed, Benchmarking Hypercube Hardware and Software, Technical Report, UIUCDCS-R-86-1303, Department of Computer Science, University of Illinois at Urbana-Champaign (1986).

  15. T. N. Mudge, G. D. Buzzard, and T. S. Abdel-Rahman, A High Performance Operating System for the NCUBE,Proc. of the 2nd Conf. on Hypercube Multiprocessors, (1986).

  16. C. T. King, Pipelined Data Parallel Algorithms—Concept, Modeling, and Design, Ph.D. Disseration, Department of Computer Science, Michigan State University (1988).

  17. J. L. Gustafson and G. R. Montry, Programming and Performance on a Cube-Connected Architecture, Proc. COMPCOM, pp. 97–100 (1988).

  18. V. Cherkassky and R. Smith, Efficient Mapping and Implementation of Matrix Algorithms on a Hypercube, Technical Report, Department of Electrical Engineering, Univ. of Minn. (1987).

  19. G. C. Fox, S. W. Otto, and A. J. Hey, Matrix Algorithms on a Hypercube I: Matrix Multiplication,Parallel Computing, pp. 17–31 (January 1987).

  20. C. T. King, W. H. Chou, and L. M. Ni, Pipelined Data Parallel Algorithms—Concept and Modeling,Proc. of the ACM Int'l Conf. on Supercomputing, (July 1988).

  21. C. Moler, Matrix Computation on Distributea Memory Multiprocessors,Proc. of First Conf. on Hypercube Multiprocessors, pp. 181–195 (1985).

Download references

Author information

Authors and Affiliations

  1. Department of Computer Science, Michigan State University, 48824-1027, East Lansing, Michigan

    Lionel M. Ni

  2. Department of Computer and Information Science, New Jersey Institute of Technology, Institute for Integrated Systems Research, 07102, Newark, New Jersey

    Chung-Ta King

Authors
  1. Lionel M. Ni
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Chung-Ta King
    View author publications

    You can also search for this author in PubMed Google Scholar

Additional information

This research was supported in part by the DARPA ACMP Project and in part by the NSF grant CCR-87-16833.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Ni, L.M., King, CT. On partitioning and mapping for hypercube computing. Int J Parallel Prog 17, 475–495 (1988). https://doi.org/10.1007/BF01407815

Download citation

  • Received:

  • Revised:

  • Issue Date:

  • DOI: https://doi.org/10.1007/BF01407815

Key Words

Search

Navigation