Parallel Cholesky method on MIMD with shared memory
Journal Article
·
· Computers and Structures
- Tire Research General Tire Inc., Akron, OH (United States)
An efficient parallel algorithm based on the Cholesky decomposition is proposed for stiffness matrix stored in a skyline form on a MIMD machine with shared memory. Included in this study is an analysis of several factors affecting the efficiency of a direct parallel solver. Based on our analysis, a multi-level synchronization check algorithm is proposed to reduce the overhead work due to data interactions among processors, thus maximizing the parallelism. In addition, a pre-ordering algorithm is implemented to avoid waiting resulting from variable column heights of the stiffness matrix. Several benchmark problems are presented to show the speedup and efficiency of the proposed algorithms. 19 refs.
- OSTI ID:
- 117661
- Journal Information:
- Computers and Structures, Journal Name: Computers and Structures Journal Issue: 1 Vol. 56; ISSN 0045-7949; ISSN CMSTCJ
- Country of Publication:
- United States
- Language:
- English
Similar Records
A supernodal Cholesky factorization algorithms for shared-memory multiprocessors
A Supernodal Cholesky Factorization Algorithm for Shared-Memory Multiprocessors
Numerical integration of hydraulic network conservation equations on MIMD parallel computers
Journal Article
·
Thu Jul 01 00:00:00 EDT 1993
· SIAM Journal on Scientific and Statistical Computing (Society for Industrial and Applied Mathematics); (United States)
·
OSTI ID:7369124
A Supernodal Cholesky Factorization Algorithm for Shared-Memory Multiprocessors
Technical Report
·
Sun Mar 31 23:00:00 EST 1991
·
OSTI ID:5896533
Numerical integration of hydraulic network conservation equations on MIMD parallel computers
Journal Article
·
Fri Mar 31 23:00:00 EST 1989
· Nuclear Science and Engineering; (USA)
·
OSTI ID:5395751