Robust parallel smoothing for multigrid via sparse approximate inverses (Q2780585)
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: Robust parallel smoothing for multigrid via sparse approximate inverses |
scientific article; zbMATH DE number 1729197
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | Robust parallel smoothing for multigrid via sparse approximate inverses |
scientific article; zbMATH DE number 1729197 |
Statements
15 April 2002
0 references
smoothing property
0 references
parallel computation
0 references
multigrid
0 references
comparison of methods
0 references
damped Jacobi method
0 references
Gauss-Seidel method
0 references
incomplete LU-factorization
0 references
sparse approximate inverses
0 references
smoothers
0 references
preconditioners
0 references
Robust parallel smoothing for multigrid via sparse approximate inverses (English)
0 references
Sparse approximate inverses (SPAI) are matrices \(M=(m_1 , m_2, \ldots , m_n)\) with a given sparsity pattern such that \(|e_k - A m_k|\) is minimal for each \(k\). Explicitly known smoothers (or preconditioners) have the advantage that they are suitable for parallel computers in contrast to Gauss-Seidel or ILU. It is shown that SPAI(0), i.e. diagonal matrices of this type are better than damped Jacobi although there is no parameter in SPAI(0). There are also some results for lesss sparse cases.
0 references
