Matrix Product on Heterogeneous Master-Worker Platforms

Abstract : This paper is focused on designing efficient parallel matrix-product algorithms for heterogeneous master-worker platforms. While matrix-product is well-understood for homogeneous 2D-arrays of processors (e.g., Cannon algorithm and ScaLAPACK outer product algorithm), there are three key hypotheses that render our work original and innovative:-Centralized data. We assume that all matrix files originate from, and must be returned to, the master. The master distributes data and computations to the workers (while in ScaLAPACK, input and output matrices are supposed to be equally distributed among participating resources beforehand). Typically, our approach is useful in the context of speeding up MATLAB or SCILAB clients running on a server (which acts as the master and initial repository of files).-Heterogeneous star-shaped platforms. We target fully heterogeneous platforms, where computational resources have different computing powers. Also, the workers are connected to the master by links of different capacities. This framework is realistic when deploying the application from the server, which is responsible for enrolling authorized resources.-Limited memory. As we investigate the parallelization of large problems, we cannot assume that full matrix column blocks can be stored in the worker memories and be re-used for subsequent updates (as in ScaLAPACK). We have devised efficient algorithms for resource selection (de-ciding which workers to enroll) and communication ordering (both for input and result messages), and we report a set of numerical experiments on a platform at our site. The experiments show that our matrix-product algorithm has smaller execution times than existing ones, while it also uses fewer resources.
Type de document :
Communication dans un congrès
13th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, 2008, Salt Lake City, Utah, United States. pp.53--62, 2008, 〈10.1145/1345206.1345217〉
Liste complète des métadonnées

https://hal.inria.fr/hal-00803487
Contributeur : Equipe Roma <>
Soumis le : mardi 20 novembre 2018 - 17:29:30
Dernière modification le : mercredi 12 décembre 2018 - 15:36:03

Fichier

ppopp_matmult.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

Collections

Citation

Jack Dongarra, Jean-François Pineau, Yves Robert, Frédéric Vivien. Matrix Product on Heterogeneous Master-Worker Platforms. 13th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, 2008, Salt Lake City, Utah, United States. pp.53--62, 2008, 〈10.1145/1345206.1345217〉. 〈hal-00803487〉

Partager

Métriques

Consultations de la notice

117

Téléchargements de fichiers

5