Partitioning sparse rectangular matrices for parallel computations of Ax and ATv

  • Bruce Hendrickson
  • Tamara G. Kolda
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1541)

Abstract

This paper addresses the problem of partitioning the nonzeros of sparse nonsymmetric and nonsquare matrices in order to efficiently compute parallel matrix-vector and matrix-transpose-vector multiplies. Our goal is to balance the work per processor while keeping communications costs low. Although the symmetric partitioning problem has been well-studied, the nonsymmetric and rectangular cases have received scant attention. We show that this problem can be described as a partitioning problem on a bipartite graph. We then describe how to use (modified) multilevel methods to partition these graphs and how to implement the matrix multiplies in parallel to take advantage of the partitioning. Finally, we compare various multilevel and other partitioning strategies on matrices from different applications. The multilevel methods are shown to be best.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Stephen T. Barnard and Horst D. Simon. A fast multilevel implementation of recursive spectral bisection for partitioning unstructured problems. Concurrency: Practice and Experience, 6:101–117, 1994.CrossRefGoogle Scholar
  2. 2.
    Michael W. Berry, Bruce Hendrickson, and Padma Raghavan. Sparse matrix reordering schemes for browsing hypertext. In James Renegar, Michael Shub, and Steve Smale, editors, The Mathematics of Numerical Analysis, volume 32 of Lectures in Applied Mathematics, pages 99–122. American Mathematical Society, 1996.Google Scholar
  3. 3.
    Roland W. Freund and Noël M. Nachtigal. QMR: A quasi-minimal residual method for non-Hermitian linear systems. Numer. Math., 60:315–339, 1991.MathSciNetCrossRefMATHGoogle Scholar
  4. 4.
    Michael R. Garey and David S. Johnson. Computers and Intractability: A Guide to the Theory of NP-Completeness. W. H. Freeman and Company, New York, 1979.MATHGoogle Scholar
  5. 5.
    Markus Hegland. Description and use of animal breeding data for large least squares problems. Technical Report TR-PA-93-50, CERFACS, Toulouse, France, 1993.Google Scholar
  6. 6.
    Bruce Hendrickson and Tamara G. Kolda. Partitioning nonsquare and nonsymmetric matrices for parallel processing. Technical Memorandum TM-13657, Oak Ridge National Laboratory, Oak Ridge, TN 37831, 1998. Submitted to SIAM J. Scientific Computing.Google Scholar
  7. 7.
    Bruce Hendrickson and Robert Leland. The Chaco user’s guide, version 2.0. Technical Report SAND95-2344, Sandia Natl. Lab., Albuquerque, NM, 87185, 1995.Google Scholar
  8. 8.
    Bruce Hendrickson and Robert Leland. A multilevel algorithm for partitioning graphs. In Proc. Supercomputing ’95. ACM, 1995.Google Scholar
  9. 9.
    A. Hofer. Schätzung von Zuchtwerten feldgeprüfter Schweine mit einem Mehrmerkmals-Tiermodell. PhD thesis, ETH-Zurich, 1990. Cited in Markus Hegland. Description and use of animal breeding data for large least squares problems. Technical Report TR-PA-93-50, CERFACS, Toulouse, France, 1993.Google Scholar
  10. 10.
    George Karypis and Vipin Kumar. A fast and high quality multilevel scheme for partitioning irregular graphs. Technical Report 95-035, Dept. Computer Science, Univ. Minnesota, Minneapolis, MN 55455, 1995.Google Scholar
  11. 11.
    George Karypis and Vipin Kumar. Parallel multilevel graph partitioning. Technical Report 95-036, Dept. Computer Science, Univ. Minnesota, Minneapolis, MN 55455, 1995.Google Scholar
  12. 12.
    B. W. Kernighan and S. Lin. An efficient heuristic procedure for partitioning graphs. Bell System Technical J., 1970.Google Scholar
  13. 13.
    Tamara G. Kolda. Partitioning sparse rectangular matrices for parallel processing. In Proc. 5th Intl. Symposium on Solving Irregularly Structured Problems in Parallel (Irregular ’98), to appear.Google Scholar
  14. 14.
    Christopher C. Paige and Michael A. Saunders. LSQR: An algorithm for sparse linear equations and sparse least squares. ACM Trans. Mathematical Software, 8:43–71, 1982.MathSciNetCrossRefMATHGoogle Scholar
  15. 15.
    Weichung Wang and Dianne P. O’Leary. Adaptive use of iterative methods in interior point methods for linear programming. Technical Report CS-TR-3560, Dept. Computer Science, Univ. Maryland, College Park, MD 20742, 1995.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1998

Authors and Affiliations

  • Bruce Hendrickson
    • 1
  • Tamara G. Kolda
    • 2
  1. 1.Parallel Computing Sciences DepartmentSandia National LabsAlbuquerque
  2. 2.Computer Science and Mathematics DivisionOak Ridge National LaboratoryOak Ridge

Personalised recommendations