Result: Multi-pass mapping schemes for parallel sparse matrix computations

Title:
Multi-pass mapping schemes for parallel sparse matrix computations
Source:
Computational science (Atlanta GA, 22-25 May 2005)Lecture notes in computer science. :245-255
Publisher Information:
Berlin: Springer, 2005.
Publication Year:
2005
Physical Description:
print, 20 ref 3
Original Material:
INIST-CNRS
Document Type:
Conference Conference Paper
File Description:
text
Language:
English
Author Affiliations:
Department of Computer Science and Engineering, The Pennsylvania State University, 343K 1ST Building, University Park, PA 16802-6106, United States
ISSN:
0302-9743
Rights:
Copyright 2005 INIST-CNRS
CC BY 4.0
Sauf mention contraire ci-dessus, le contenu de cette notice bibliographique peut être utilisé dans le cadre d’une licence CC BY 4.0 Inist-CNRS / Unless otherwise stated above, the content of this bibliographic record may be used under a CC BY 4.0 licence by Inist-CNRS / A menos que se haya señalado antes, el contenido de este registro bibliográfico puede ser utilizado al amparo de una licencia CC BY 4.0 Inist-CNRS
Notes:
Computer science; theoretical automation; systems
Accession Number:
edscal.16991697
Database:
PASCAL Archive

Further Information

Consider the solution of a large sparse linear system Ax = b on multiprocessors. A parallel sparse matrix factorization is required in a direct solver. Alternatively, if Krylov subspace iterative methods are used, then incomplete forms of parallel sparse factorization are required for preconditioning. In such schemes, the underlying parallel computation is tree-structured, utilizing task-parallelism at lower levels of the tree and data-parallelism at higher levels. The proportional heuristic has typically been used to map the data and computation to processors. However, for sparse systems from finite-element methods on complex domains, the resulting assignments can exhibit significant load-imbalances. In this paper, we develop a multi-pass mapping scheme to reduce such load imbalances and we demonstrate its effectiveness for a test suite of large sparse matrices. Our scheme can also be used to generate improved mappings for tree-structured applications beyond those considered in this paper.