You need JavaScript to view this

Parallelization of applications for networks with homogeneous and heterogeneous processors; Parallelisation d`applications pour des reseaux de processeurs homogenes ou heterogenes

Abstract

The aim of this thesis is to study and develop efficient methods for parallelization of scientific applications on parallel computers with distributed memory. The first part presents two libraries of PVM (Parallel Virtual Machine) and MPI (Message Passing Interface) communication tools. They allow implementation of programs on most parallel machines, but also on heterogeneous computer networks. This chapter illustrates the problems faced when trying to evaluate performances of networks with heterogeneous processors. To evaluate such performances, the concepts of speed-up and efficiency have been modified and adapted to account for heterogeneity. The second part deals with a study of parallel application libraries such as ScaLAPACK and with the development of communication masking techniques. The general concept is based on communication anticipation, in particular by pipelining message sending operations. Experimental results on Cray T3D and IBM SP1 machines validates the theoretical studies performed on basic algorithms of the libraries discussed above. Two examples of scientific applications are given: the first is a model of young stars for astrophysics and the other is a model of photon trajectories in the Compton effect. (J.S.). 83 refs., 65 figs., 24 tabs.
Authors:
Publication Date:
Oct 07, 1994
Product Type:
Thesis/Dissertation
Report Number:
FRNC-TH-3757
Reference Number:
SCA: 990200; 550601; PA: AIX-27:048911; EDB-96:103330; NTS-96:018753; SN: 96001612151
Resource Relation:
Other Information: TH: These (D. es Sc.).; PBD: 7 Oct 1994
Subject:
99 MATHEMATICS, COMPUTERS, INFORMATION SCIENCE, MANAGEMENT, LAW, MISCELLANEOUS; 55 BIOLOGY AND MEDICINE, BASIC STUDIES; PARALLEL PROCESSING; COMPUTERIZED SIMULATION; PERFORMANCE TESTING; ASTRONOMY; COMMUNICATIONS; COMPTON EFFECT; COMPTON SCATTERING TOMOGRAPHY; COMPUTER ARCHITECTURE; COMPUTER NETWORKS; COSMIC PHOTONS; CRAY COMPUTERS; DATA BASE MANAGEMENT; DYNAMIC PROGRAMMING; EQUIPMENT INTERFACES; IBM COMPUTERS; LAPLACIAN; LIBRARIES; MEMORY MANAGEMENT; MONTE CARLO METHOD; PHOTON-ELECTRON INTERACTIONS; PROGRAMMING LANGUAGES; SEQUENTIAL CIRCUITS; STAR MODELS; SUPERCOMPUTERS; TASK SCHEDULING; TIME DELAY; TRAJECTORIES
OSTI ID:
246829
Research Organizations:
Institut National Polytechnique, 38 - Grenoble (France)
Country of Origin:
France
Language:
French
Other Identifying Numbers:
Other: ON: DE96627724; TRN: FR9600920048911
Availability:
INIS; OSTI as DE96627724
Submitting Site:
FRN
Size:
152 p.
Announcement Date:
Jul 12, 1996

Citation Formats

Colombet, L. Parallelization of applications for networks with homogeneous and heterogeneous processors; Parallelisation d`applications pour des reseaux de processeurs homogenes ou heterogenes. France: N. p., 1994. Web.
Colombet, L. Parallelization of applications for networks with homogeneous and heterogeneous processors; Parallelisation d`applications pour des reseaux de processeurs homogenes ou heterogenes. France.
Colombet, L. 1994. "Parallelization of applications for networks with homogeneous and heterogeneous processors; Parallelisation d`applications pour des reseaux de processeurs homogenes ou heterogenes." France.
@misc{etde_246829,
title = {Parallelization of applications for networks with homogeneous and heterogeneous processors; Parallelisation d`applications pour des reseaux de processeurs homogenes ou heterogenes}
author = {Colombet, L}
abstractNote = {The aim of this thesis is to study and develop efficient methods for parallelization of scientific applications on parallel computers with distributed memory. The first part presents two libraries of PVM (Parallel Virtual Machine) and MPI (Message Passing Interface) communication tools. They allow implementation of programs on most parallel machines, but also on heterogeneous computer networks. This chapter illustrates the problems faced when trying to evaluate performances of networks with heterogeneous processors. To evaluate such performances, the concepts of speed-up and efficiency have been modified and adapted to account for heterogeneity. The second part deals with a study of parallel application libraries such as ScaLAPACK and with the development of communication masking techniques. The general concept is based on communication anticipation, in particular by pipelining message sending operations. Experimental results on Cray T3D and IBM SP1 machines validates the theoretical studies performed on basic algorithms of the libraries discussed above. Two examples of scientific applications are given: the first is a model of young stars for astrophysics and the other is a model of photon trajectories in the Compton effect. (J.S.). 83 refs., 65 figs., 24 tabs.}
place = {France}
year = {1994}
month = {Oct}
}