Block cyclic array redistribution

Abstract : Implementing linear algebra kernels on distributed memory parallel computers raises the problem of data distribution of matrices and vectors among the processors. Block-cyclic distribution seems to suit well for most algorithms. But one has to choose a good compromise for the size of the blocks (to achieve a good efficiency and a good load balancing). This choice heavily depends on each operation, so it is essential to be able to go from one distribution to another very quickly. We present here the algorithms we implemented in the SCALAPACK library. A complexity study is then made that proves the efficiency of our solution. Timing results on a network of SUN workstations and the Cray T3D using PVM corroborates the results.
Document type :
Reports
Complete list of metadatas

Cited literature [11 references]  Display  Hide  Download

https://hal-lara.archives-ouvertes.fr/hal-02101963
Contributor : Colette Orange <>
Submitted on : Wednesday, April 17, 2019 - 9:10:51 AM
Last modification on : Friday, May 17, 2019 - 1:39:19 AM

File

RR1995-39.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02101963, version 1

Collections

Citation

Loïc Prylli, Bernard Tourancheau. Block cyclic array redistribution. [Research Report] LIP RR-1995-39, Laboratoire de l'informatique du parallélisme. 1995, 2+12p. ⟨hal-02101963⟩

Share

Metrics

Record views

2

Files downloads

7