Parallel Linear Algebra


Linear algebra calculations are at the core of most scientific computations. For this reason, efficient parallel algorithms for linear algebra operations are critical to the effective use of parallel machines. I have worked on the parallelization of direct linear- and eigen-solvers as well iterative methods. In addition, I have worked on parallel algorithms for a variety of scientific computing problems which don't involve linear algebra as described here .


Papers

  • Towards an Efficient Parallel Eigensolver for Dense Symmetric Matrices, Bruce Hendrickson, Elizabeth Jessup and Christopher Smith. SIAM J. Sci. Comput. 20(1):1132-1154, 1999.
    Paper, Abstract
  • A Parallel Algorithm for Householder Tridiagonalization, Christopher Smith, Bruce Hendrickson and Elizabeth Jessup. In Proc. 5th SIAM Conf. Appl. Lin. Alg., 1994.
    Paper, Abstract
  • An Efficient Parallel Algorithm for Matrix-Vector Multiplication, Bruce Hendrickson, Robert Leland and Steve Plimpton. Intl. J. High Speed Comput., 7(1):73-88, 1995.
    Paper, Abstract
  • The Torus-Wrap Mapping for Dense Matrix Calculations on Massively Parallel Computers, Bruce Hendrickson and David Womble. SIAM J. Sci. Stat. Comput., 15(5):1201-1226, 1994.
    Paper, Abstract

  • Return to Bruce's home page