Parallel numerical asynchronous iterative algorithms: Large scale experimentations

Author

Charr, Jean-Claude ; Couturier, Raphaël ; Laiymani, David

Author_Institution

Lab. of Comput. Sci., Univ. of Franche-Comte (LIFC), Belfort, France

fYear

2009

fDate

23-29 May 2009

Firstpage

1

Lastpage

8

Abstract

This paper presents many typical problems that are encountered when executing large scale scientific applications over distributed architectures. The causes and effects of these problems are explained and a solution for some classes of scientific applications is also proposed. This solution is the combination of the asynchronous iteration model with JACEP2P-V2 which is a fully decentralized and fault tolerant platform dedicated to executing parallel asynchronous applications over volatile distributed architectures. We explain in detail how our approach deals with each of these problems. Then we present two large scale numerical experiments that prove the efficiency and the robustness of our approach.

Keywords

distributed processing; iterative methods; natural sciences computing; parallel algorithms; software architecture; JACEP2P-V2; asynchronous iteration model; fault tolerant platform; fully decentralized platform; large scale experimentation; large scale scientific application; parallel asynchronous application; parallel numerical asynchronous iterative algorithm; volatile distributed architecture; Computational modeling; Computer architecture; Computer crashes; Concurrent computing; Convergence; Distributed computing; Iterative algorithms; Iterative methods; Laboratories; Large-scale systems; Large scale computing; parallel asynchronous iterative algorithms; volatile environment;

fLanguage

English

Publisher

ieee

Conference_Titel

Parallel & Distributed Processing, 2009. IPDPS 2009. IEEE International Symposium on

Conference_Location

Rome

ISSN

1530-2075

Print_ISBN

978-1-4244-3751-1

Electronic_ISBN

1530-2075

Type

conf

DOI

10.1109/IPDPS.2009.5161158

Filename

5161158