Mapping Asynchronous Iterative Applications on Heterogeneous Distributed Architectures

Abstract : To design parallel numerical algorithms on large scale distributed and heterogeneous platforms, the asynchronous iteration model (AIAC) may be an efficient solution. This class of algorithm is very suitable since it enables communication/computation overlapping and it suppresses all synchronizations between computation nodes. Since target architectures are composed of more than one thousand heterogeneous nodes connected through heterogeneous networks, the need for mapping algorithms is crucial. In this paper, we propose a new mapping algorithm dedicated to the AIAC model. To evaluate our mapping algorithm we implemented it in the JaceP2P programming and executing environment dedicated to AIAC applications and we conducted a set of experiments on the Grid'5000 testbed. Results are very encouraging and show that the use of our algorithm brings an important gain in term of execution time (about 40%).
Type de document :
Communication dans un congrès
IPDPS'10, ACM/IEEE Int. Parallel and Distributed Processing Symposium, Workshop on Parallel and Distributed Scientific and Engineering Computing, Apr 2010, Atlanta, United States. 2010, 〈10.1109/IPDPSW.2010.5470762〉
Liste complète des métadonnées

https://hal.inria.fr/hal-00684533
Contributeur : Ist Rennes <>
Soumis le : lundi 2 avril 2012 - 13:58:26
Dernière modification le : jeudi 11 janvier 2018 - 06:16:36

Identifiants

Citation

Raphaël Couturier, David Laiymani, Sébastien Miquée. Mapping Asynchronous Iterative Applications on Heterogeneous Distributed Architectures. IPDPS'10, ACM/IEEE Int. Parallel and Distributed Processing Symposium, Workshop on Parallel and Distributed Scientific and Engineering Computing, Apr 2010, Atlanta, United States. 2010, 〈10.1109/IPDPSW.2010.5470762〉. 〈hal-00684533〉

Partager

Métriques

Consultations de la notice

97