Abstract | ||
---|---|---|
:It is demonstrated how conventional algorithms for computingthe LDU decomposition of a square matrix, or Choleskyfactorization for symmetric positive definite matrices, can bere-ordered into Jacobi-type algorithms. For efficient parallelimplementation on a systolic array, the resulting schemescompare favorably with earlier implementations.List of symbols :Matrices are denoted by capitals A, X, etc., sometimes with an iterationindex X 1 , X 2 , etc.X i;j denotes the (i; j)... |
Year | DOI | Venue |
---|---|---|
1991 | 10.1016/0743-7315(91)90114-O | J. Parallel Distrib. Comput. |
Keywords | Field | DocType |
jacobi-type algorithm,cholesky factorization,systolic array | Algebra,Incomplete Cholesky factorization,Computer science,Matrix (mathematics),Matrix decomposition,Minimum degree algorithm,Positive-definite matrix,Square matrix,Algorithm,Factorization,Cholesky decomposition | Journal |
Volume | Issue | ISSN |
13 | 1 | Journal of Parallel and Distributed Computing |
Citations | PageRank | References |
1 | 0.63 | 0 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Marc Moonen | 1 | 377 | 46.79 |
Paul van Dooren | 2 | 649 | 90.48 |
Joos Vandewalle | 3 | 4420 | 523.42 |