Coarse-Grained Parallel Matrix-Free Solution of a Three-Dimensional Elliptic Prototype Problem

  • Kevin P. Allen
  • Matthias K. Gobbert
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2668)


The finite difference discretization of the Poisson equation in three dimensions results in a large, sparse, and highly structured system of linear equations. This prototype problem is used to analyze the performance of the parallel linear solver on coarse-grained clusters of workstations. The conjugate gradient method with a matrix-free implementation of the matrix-vector product with the system matrix is shown to be optimal with respect to memory usage and runtime performance. Parallel performance studies con.rm that speedup can be obtained. When only an ethernet interconnect is available, best performance is limited to up to 4 processors, since the conjugate gradient method involves several communications per iteration. Using a high performance Myrinet interconnect, excellent speedup is possible for at least up to 32 processors. These results justify the use of this linear solver as the computational kernel for the time-stepping in a system of reaction-diffusion equations.


Conjugate Gradient Method System Matrix Memory Usage Biophysical Journal Computational Kernel 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    K. P. Allen and M. K. Gobbert. A matrix-free conjugate gradient method for cluster computing. Submitted.Google Scholar
  2. 2.
    J. W. Demmel. Applied Numerical Linear Algebra. SIAM, 1997.Google Scholar
  3. 3.
    W. Gropp, E. Lusk, and A. Skjellum. Using MPI: Portable Parallel Programming with the Message-Passing Interface. MIT Press, second edition, 1999.Google Scholar
  4. 4.
    C. A. Hall and T. A. Porsching. Numerical Analysis of Partial Differential Equations. Prentice-Hall, 1990.Google Scholar
  5. 5.
    A. L. Hanhart. Coarse-grained parallel solution of a three-dimensional model for calcium concentration in human heart cells. M.S. thesis, University of Maryland, Baltimore County, 2002.Google Scholar
  6. 6.
    A. L. Hanhart, M. K. Gobbert, and L. T. Izu. A memory-efficient finite element method for a calcium concentration model in human heart cells. Submitted.Google Scholar
  7. 7.
    A. Iserles. A First Course in the Numerical Analysis of Differential Equations. Cambridge Texts in Applied Mathematics. Cambridge University Press, 1996.Google Scholar
  8. 8.
    L. T. Izu, J. R. H. Mauban, C. W. Balke, and W. G. Wier. Large currents generate cardiac Ca2+ sparks. Biophysical Journal, 80:88–102, 2001.Google Scholar
  9. 9.
    L. T. Izu, W. G. Wier, and C. W. Balke. Theoretical analysis of the Ca2+ spark amplitude distribution. Biophysical Journal, 75:1144–1162, 1998.Google Scholar
  10. 10.
    L. T. Izu, W. G. Wier, and C. W. Balke. Evolution of cardiac calcium waves from stochastic calcium sparks. Biophysical Journal, 80:103–120, 2001.CrossRefGoogle Scholar
  11. 11.
    D. S. Watkins. Fundamentals of Matrix Computations. Wiley, second edition, 2002.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2003

Authors and Affiliations

  • Kevin P. Allen
    • 1
  • Matthias K. Gobbert
    • 1
  1. 1.Department of Mathematics and StatisticsUniversity of Maryland, Baltimore CountyBaltimoreUSA

Personalised recommendations