Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

A scalable 2-D parallel sparse solver

Conference ·
OSTI ID:125546
;  [1]
  1. Iowa State Univ., Ames, IA (United States)

Scalability beyond a small number of processors, typically 32 or less, is known to be a problem for existing parallel general sparse (PGS) direct solvers. This paper presents a parallel general sparse PGS direct solver for general sparse linear systems on distributed memory machines. The algorithm is based on the well-known sequential sparse algorithm Y12M. To achieve efficient parallelization, a 2-D scattered decomposition of the sparse matrix is used. The proposed algorithm is more scalable than existing parallel sparse direct solvers. Its scalability is evaluated on a 256 processor nCUBE2s machine using Boeing/Harwell benchmark matrices.

OSTI ID:
125546
Report Number(s):
CONF-950212--
Country of Publication:
United States
Language:
English

Similar Records

A scalable parallel algorithm for sparse Cholesky factorization
Book · Fri Dec 30 23:00:00 EST 1994 · OSTI ID:87680

A distributed-memory hierarchical solver for general sparse linear systems
Journal Article · Tue Dec 19 23:00:00 EST 2017 · Parallel Computing · OSTI ID:1429626

SuperLU{_}DIST: A scalable distributed-memory sparse direct solver for unsymmetric linear systems
Journal Article · Tue Mar 26 23:00:00 EST 2002 · ACM Transaction on Mathematical Software · OSTI ID:836786