Parallel compact multi-dimensional numerical algorithm with application to aeroacoustics

Alex Povitsky, Philip John Morris

Research output: Contribution to conferencePaper

4 Citations (Scopus)

Abstract

In this study we propose an efficient method to parallelize high-order compact numerical algorithms for the solution of three-dimensional PDEs in a spacetime domain. The proposed parallelization method includes 3-D partitioning and the solution of sets of linear banded systems that split onto processors. Efficient. parallelization of the Thomas algorithm is difficult due to its pipelined nature. We use processors for either non-local data independent computations, solving lines in the next spatial direction, or local datadependent computations by the Runge-Kutta method while processors are idle due to the forward and backward recurrences of the Thomas algorithm. Control of processor communication and computations by a static schedule is adopted in this study. The obtained parallelization speed-up of the novel algorithm is about twice as much as that for the standard pipelined algorithm and close to that for the explicit algorithm.

Original languageEnglish (US)
StatePublished - Jan 1 1999
Event14th Computational Fluid Dynamics Conference, 1999 - Norfolk, United States
Duration: Nov 1 1999Nov 5 1999

Other

Other14th Computational Fluid Dynamics Conference, 1999
CountryUnited States
CityNorfolk
Period11/1/9911/5/99

Fingerprint

Aeroacoustics
Runge Kutta methods
Linear systems
Communication

All Science Journal Classification (ASJC) codes

  • Engineering(all)

Cite this

Povitsky, A., & Morris, P. J. (1999). Parallel compact multi-dimensional numerical algorithm with application to aeroacoustics. Paper presented at 14th Computational Fluid Dynamics Conference, 1999, Norfolk, United States.
Povitsky, Alex ; Morris, Philip John. / Parallel compact multi-dimensional numerical algorithm with application to aeroacoustics. Paper presented at 14th Computational Fluid Dynamics Conference, 1999, Norfolk, United States.
@conference{03618642636d4804a3e1e6f701cdd395,
title = "Parallel compact multi-dimensional numerical algorithm with application to aeroacoustics",
abstract = "In this study we propose an efficient method to parallelize high-order compact numerical algorithms for the solution of three-dimensional PDEs in a spacetime domain. The proposed parallelization method includes 3-D partitioning and the solution of sets of linear banded systems that split onto processors. Efficient. parallelization of the Thomas algorithm is difficult due to its pipelined nature. We use processors for either non-local data independent computations, solving lines in the next spatial direction, or local datadependent computations by the Runge-Kutta method while processors are idle due to the forward and backward recurrences of the Thomas algorithm. Control of processor communication and computations by a static schedule is adopted in this study. The obtained parallelization speed-up of the novel algorithm is about twice as much as that for the standard pipelined algorithm and close to that for the explicit algorithm.",
author = "Alex Povitsky and Morris, {Philip John}",
year = "1999",
month = "1",
day = "1",
language = "English (US)",
note = "14th Computational Fluid Dynamics Conference, 1999 ; Conference date: 01-11-1999 Through 05-11-1999",

}

Povitsky, A & Morris, PJ 1999, 'Parallel compact multi-dimensional numerical algorithm with application to aeroacoustics' Paper presented at 14th Computational Fluid Dynamics Conference, 1999, Norfolk, United States, 11/1/99 - 11/5/99, .

Parallel compact multi-dimensional numerical algorithm with application to aeroacoustics. / Povitsky, Alex; Morris, Philip John.

1999. Paper presented at 14th Computational Fluid Dynamics Conference, 1999, Norfolk, United States.

Research output: Contribution to conferencePaper

TY - CONF

T1 - Parallel compact multi-dimensional numerical algorithm with application to aeroacoustics

AU - Povitsky, Alex

AU - Morris, Philip John

PY - 1999/1/1

Y1 - 1999/1/1

N2 - In this study we propose an efficient method to parallelize high-order compact numerical algorithms for the solution of three-dimensional PDEs in a spacetime domain. The proposed parallelization method includes 3-D partitioning and the solution of sets of linear banded systems that split onto processors. Efficient. parallelization of the Thomas algorithm is difficult due to its pipelined nature. We use processors for either non-local data independent computations, solving lines in the next spatial direction, or local datadependent computations by the Runge-Kutta method while processors are idle due to the forward and backward recurrences of the Thomas algorithm. Control of processor communication and computations by a static schedule is adopted in this study. The obtained parallelization speed-up of the novel algorithm is about twice as much as that for the standard pipelined algorithm and close to that for the explicit algorithm.

AB - In this study we propose an efficient method to parallelize high-order compact numerical algorithms for the solution of three-dimensional PDEs in a spacetime domain. The proposed parallelization method includes 3-D partitioning and the solution of sets of linear banded systems that split onto processors. Efficient. parallelization of the Thomas algorithm is difficult due to its pipelined nature. We use processors for either non-local data independent computations, solving lines in the next spatial direction, or local datadependent computations by the Runge-Kutta method while processors are idle due to the forward and backward recurrences of the Thomas algorithm. Control of processor communication and computations by a static schedule is adopted in this study. The obtained parallelization speed-up of the novel algorithm is about twice as much as that for the standard pipelined algorithm and close to that for the explicit algorithm.

UR - http://www.scopus.com/inward/record.url?scp=84983120200&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84983120200&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:84983120200

ER -

Povitsky A, Morris PJ. Parallel compact multi-dimensional numerical algorithm with application to aeroacoustics. 1999. Paper presented at 14th Computational Fluid Dynamics Conference, 1999, Norfolk, United States.