Data Allocation Strategies for Dense Linear Algebra Kernels on Heterogeneous Two-dimensional Grids

Abstract : We study the implementation of dense linear algebra computations, such as matrix multiplication and linear system solvers, on two-dimensional (2D) grids of heterogeneous processors. For these operations, 2D-grids are the key to scalability and efficiency. The uniform block-cyclic data distribution scheme commonly used for homogeneous collections of processors limits the performance of these operations on heterogeneous grids to the speed of the slowest processor. We present and study more sophisticated data allocation strategies that balance the load on heterogeneous 2D-grids with respect to the performance of the processors. The practical usefulness of these strategies is fully demonstrated by experimental data for a heterogeneous network of workstations.
Document type :
Reports
Complete list of metadatas

https://hal-lara.archives-ouvertes.fr/hal-02102033
Contributor : Colette Orange <>
Submitted on : Wednesday, April 17, 2019 - 9:12:30 AM
Last modification on : Sunday, April 28, 2019 - 1:23:06 AM

File

RR1999-31.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02102033, version 1

Collections

Citation

Vincent Boudet, Antoine Petitet, Fabrice Rastello, Yves Robert. Data Allocation Strategies for Dense Linear Algebra Kernels on Heterogeneous Two-dimensional Grids. [Research Report] LIP RR-1999-31, Laboratoire de l'informatique du parallélisme. 1999, 2+24 p. ⟨hal-02102033⟩

Share

Metrics

Record views

10

Files downloads

19