Efficient parallel algorithms for dense Cholesky factorization (Q4249175)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 1302627
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Efficient parallel algorithms for dense Cholesky factorization |
scientific article; zbMATH DE number 1302627 |
Statements
13 September 1999
0 references
dense Cholesky factorization
0 references
parallel algorithms
0 references
LogP
0 references
latency
0 references
overhead
0 references
gap
0 references
number of processor/memory pairs
0 references
0.8088274002075195
0 references
0.8078706860542297
0 references
0.8031956553459167
0 references
Efficient parallel algorithms for dense Cholesky factorization (English)
0 references
Dense Cholesky factorization arises in areas such as linear programming, boundary element methods and radiosity computations. In this paper we focus on designing efficient parallel algorithms for Cholesky factorization. In order to take portability into account, we utilize LogP in our performance analysis. LogP is a general parallel model utilizing point-to-point communication. The model captures important characteristics of parallel machines without taking into account specifics such as the interconnection network in order to deal with portability issues. The four parameters of LogP are: \(({\mathbf L})\): latency; \(({\mathbf o})\): overhead; \(({\mathbf g})\): gap; and \(({\mathbf P})\): number of processor/memory pairs. For ease of discussion, we assume that \(L= g\) and \(o= 0\). For the entire collection see [Zbl 0909.00047].
0 references