Parallel factorizations and parallel solvers for tridiagonal linear systems (Q1194530)
From MaRDI portal
 | This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use this page instead for the normal view: Parallel factorizations and parallel solvers for tridiagonal linear systems |
scientific article; zbMATH DE number 64496
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Parallel factorizations and parallel solvers for tridiagonal linear systems |
scientific article; zbMATH DE number 64496 |
Statements
Parallel factorizations and parallel solvers for tridiagonal linear systems (English)
0 references
27 September 1992
0 references
At first a unified representation of all types of parallel methods (splittings) for tridiagonal systems is presented. Then parallel algorithms for LU, Gauss-Jordan and cyclic reduction are investigated on a 32-transputer system. The last one is the best one. The speedup in the paper is not clearly defined. It should be the comparison to the best scalar algorithm with roughly half the number of operations of the parallel algorithm. It should be mentioned that for the same reason many tridiagonal systems (occurring for ADI or time split methods) are better solved in parallel, each system on a single processor.
0 references
Gauss-Jordan matrix
0 references
LU-factorization
0 references
parallel methods
0 references
splittings
0 references
tridiagonal systems
0 references
parallel algorithms
0 references
cyclic reduction
0 references
0 references
0 references
0 references
0.9158264994621276
0 references
0.9158264994621276
0 references
0.8851321339607239
0 references
0.8699362277984619
0 references
