Impact of mixed-parallelism on parallel implementations of the Strassen and Winograd matrix multiplication algorithms

F. Desprez, F. Suter

Research output: Contribution to journalArticlepeer-review

22 Scopus citations

Abstract

In this paper we study the impact of the simultaneous exploitation of data- and task-parallelism, so called mixed-parallelism, on the Strassen and Winograd matrix multiplication algorithms. This work takes place in the context of Grid computing and, in particular, in the Client-Agent(s)-Server(s) model, where data can already be distributed on the platform. For each of those algorithms, we propose two mixed-parallel implementations. The former follows the phases of the original algorithms while the latter has been designed as the result of a list scheduling algorithm. We give a theoretical comparison, in terms of memory usage and execution time, between our algorithms and classical data-parallel implementations. This analysis is corroborated by experiments. Finally, we give some hints about heterogeneous and recursive versions of our algorithms.

Original languageEnglish
Pages (from-to)771-797
Number of pages27
JournalConcurrency and Computation: Practice and Experience
Volume16
Issue number8
DOIs
StatePublished - Jul 2004
Externally publishedYes

Keywords

  • Grid computing
  • Mixed-parallelism
  • ScaLAPACK
  • Strassen
  • Winograd

Fingerprint

Dive into the research topics of 'Impact of mixed-parallelism on parallel implementations of the Strassen and Winograd matrix multiplication algorithms'. Together they form a unique fingerprint.

Cite this