DMCA
Parallelizing Particle-In-Cell Codes with OpenMP and MPI (2007)
BibTeX
@MISC{Larsgård07parallelizingparticle-in-cell,
author = {Nils Magnus Larsgård and Supervisor Anne and Cathrine Elster},
title = {Parallelizing Particle-In-Cell Codes with OpenMP and MPI},
year = {2007}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
This thesis searches for the best configuration of OpenMP/MPI for optimal performance. We will run a parallel simulation-application on a modern supercomputer to measure the effects of different configurations of the mixed code. After analyzing the performance of different configurations from a hardware point of view, we will propose a general model for estimating overhead in OpenMP loops. We will parallelize a physics simulation to do large scale simulations as efficient as possible. We will look at typical physics simulation codes to parallelize and optimize them with OpenMP and MPI in a mixed mode. The goal of the parallelization is to make the code highly efficient for a parallel system and to make it as scalable as possible to do large-scale simulations as well as using this parallelized application as a benchmark for our MPI/OpenMP tests.
Keyphrases
particle-in-cell code different configuration parallelized application openmp loop typical physic simulation code modern supercomputer mixed mode large scale simulation large-scale simulation hardware point parallel system parallel simulation-application mixed code openmp mpi general model optimal performance physic simulation mpi openmp test
