Methods for Partitioning Data to Improve Parallel Execution Time for Sorting on Heterogeneous Clusters - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2006

Methods for Partitioning Data to Improve Parallel Execution Time for Sorting on Heterogeneous Clusters

Résumé

The aim of the paper is to introduce general techniques in order to optimize the parallel execution time of sorting on a distributed architectures with processors of various speeds. Such an application requires a partitioning step. For uniformly related processors (processors speeds are related by a constant factor), we develop a constant time technique for mastering processor load and execution time in an heterogeneous environment and also a technique to deal with unknown cost functions. For non uniformly related processors, we use a technique based on dynamic programming. Most of the time, the solutions are in O(p) (p is the number of processors), independent of the problem size n. Consequently, there is a small overhead regarding the problem we deal with but it is inherently limited by the knowing of time complexity of the portion of code following the partitioning.
Fichier principal
Vignette du fichier
gpc2006final.pdf (236.03 Ko) Télécharger le fichier
gpc2006.pdf (993.33 Ko) Télécharger le fichier
Format : Autre

Dates et versions

hal-00084822 , version 1 (10-07-2006)

Identifiants

Citer

Christophe Cérin, Jean-Christophe Dubacq, Jean-Louis Roch. Methods for Partitioning Data to Improve Parallel Execution Time for Sorting on Heterogeneous Clusters. GPC 2006, 2006, France. pp.175-186. ⟨hal-00084822⟩
197 Consultations
160 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More