Jump to ContentJump to Main Navigation
Parallel Scientific Computation$
Users without a subscription are not able to see the full content.

Rob H. Bisseling

Print publication date: 2004

Print ISBN-13: 9780198529392

Published to Oxford Scholarship Online: September 2007

DOI: 10.1093/acprof:oso/9780198529392.001.0001

SPARSE MATRIX–VECTOR MULTIPLICATION

Chapter:
(p. 163 ) 4 SPARSE MATRIX–VECTOR MULTIPLICATION
Source:
Parallel Scientific Computation
Author(s):

Rob H. Bisseling

Publisher:
Oxford University Press
DOI:10.1093/acprof:oso/9780198529392.003.0004

This chapter introduces irregular algorithms and presents the example of multiplying a sparse matrix with a vector, which is the central operation in iterative solvers for linear systems and eigensystems. The irregular sparsity pattern of the matrix does not change during the multiplication, and the multiplication may be repeated many times with the same matrix. This justifies putting a lot of effort in finding good data distribution for a parallel multiplication. A useful non-Cartesian matrix distribution called the Mondriaan distribution is introduced, and an algorithm for finding such a distribution for a general sparse matrix is studied. Several special types of matrices are analyzed, such as random sparse matrices and Laplacian matrices. The program of this chapter demonstrates the use of the bulk synchronous message passing primitives from BSPlib, which were designed to facilitate irregular computations.

Keywords:   irregular, iterative solver, Cartesian distribution, dense matrix-vector multiplication, Laplacian matrix, Mondriaan distribution, random sparse matrix, sparse matrix

Oxford Scholarship Online requires a subscription or purchase to access the full text of books within the service. Public users can however freely search the site and view the abstracts and keywords for each book and chapter.

Please, subscribe or login to access full text content.

If you think you should have access to this title, please contact your librarian.

To troubleshoot, please check our FAQs , and if you can't find the answer there, please contact us .