Algorithmic Issues for (Distributed) Heterogeneous Computing Platforms. Extended Abstract

Abstract : Future computing platforms will be distributed and heterogeneous. Such platforms range from heterogeneous networks of workstations (NOWs) to collections of NOWs and parallel servers scattered throughout the world and linked through high-speed networks. Implementing tightly-coupled algorithms on such platforms raises several challenging issues. New data distribution and load balancing strategies are required to squeeze the most out of heterogeneous platforms. In this paper, we first summarize previous results obtained for heterogeneous NOWs, dealing with the implementation of standard numerical kernels such as finite-difference stencils or dense linear solvers. Next we target distributed collections of heterogeneous NOWs, and we discuss data allocation strategies for dense linear solvers on top of such platforms. These results indicate that a major algorithmic and software effort is needed to come up with efficient numerical libraries on the computational grid.
Document type :
Reports
Complete list of metadatas

https://hal-lara.archives-ouvertes.fr/hal-02101801
Contributor : Colette Orange <>
Submitted on : Wednesday, April 17, 2019 - 9:06:38 AM
Last modification on : Wednesday, May 22, 2019 - 1:32:15 AM

File

RR1999-19.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02101801, version 1

Collections

Citation

Vincent Boudet, Fabrice Rastello, Yves Robert. Algorithmic Issues for (Distributed) Heterogeneous Computing Platforms. Extended Abstract. [Research Report] LIP RR-1999-19, Laboratoire de l'informatique du parallélisme. 1999, 2+9p. ⟨hal-02101801⟩

Share

Metrics

Record views

11

Files downloads

10