DocumentCode :
2162506
Title :
Performance measurement and comparison of a set of parallel periodic and non-periodic tridiagonal solvers
Author :
Sun, Xian-He ; Moitra, Stuti
Author_Institution :
Dept. of Comput. Sci., Louisiana State Univ., Baton Rouge, LA, USA
fYear :
1996
fDate :
12-14 Jun 1996
Firstpage :
480
Lastpage :
486
Abstract :
Various traditional solvers have been proposed in recent years for different parallel platforms. In this paper, the performance of three tridiagonal solvers, namely, the parallel partition LU algorithm, the parallel diagonal dominant algorithm, and the reduced diagonal dominant algorithm, is studied. These algorithms are designed for distributed-memory machines and are tested on an Intel Paragon and an IBM SP2 machines. Measured results are reported in terms of execution time and speedup. The measured results match analytical results closely. In addition to address implementation issues, performance considerations such as problem sizes and models of speedup are also discussed
Keywords :
computational complexity; distributed memory systems; matrix algebra; parallel algorithms; performance evaluation; IBM SP2; Intel Paragon; distributed-memory machines; parallel diagonal dominant algorithm; parallel partition LU algorithm; performance; problem sizes; reduced diagonal dominant algorithm; speedup; tridiagonal solvers; Algorithm design and analysis; Computer science; Kernel; Linear systems; NASA; Partitioning algorithms; Sun; Testing; Time measurement; Velocity measurement;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Parallel Architectures, Algorithms, and Networks, 1996. Proceedings., Second International Symposium on
Conference_Location :
Beijing
ISSN :
1087-4089
Print_ISBN :
0-8186-7460-1
Type :
conf
DOI :
10.1109/ISPAN.1996.509029
Filename :
509029
Link To Document :
بازگشت