A spectral element solution of the shallow-water equations on multiprocessor computers

Enrique N. Curchitser, Mohamed Iskandarani, Dale B. Haidvogel

Research output: Contribution to journalArticle

16 Citations (Scopus)

Abstract

A shallow-water spectral element ocean model is implemented on multiple instruction multiple data, distributed memory parallel computers. A communications-minimizing partitioning algorithm for unstructured meshes, based on graph theory, is presented and is shown to improve the efficiency in a limited range of granularities. A domain decomposition implementation with an architecture-independent communications scheme, using message passing, is devised and tested on an nCUBE/2, a Cray T3D, and an IBM SP2. The implementation exhibits high efficiencies over a wide range of granularities. An order of magnitude analysis shows that, to leading order, the efficiency stays constant when KN2 grows proportionally to P, where K is the total number of elements, N is the order of the spectral truncation within an element, and P is the number of processors.

Original languageEnglish (US)
Pages (from-to)510-521
Number of pages12
JournalJournal of Atmospheric and Oceanic Technology
Volume15
Issue number2
StatePublished - Apr 1998
Externally publishedYes

Fingerprint

shallow-water equation
communication
Water
shallow water
partitioning
Graph theory
Communication
Message passing
decomposition
ocean
Decomposition
Data storage equipment
analysis

ASJC Scopus subject areas

  • Atmospheric Science
  • Ocean Engineering

Cite this

A spectral element solution of the shallow-water equations on multiprocessor computers. / Curchitser, Enrique N.; Iskandarani, Mohamed; Haidvogel, Dale B.

In: Journal of Atmospheric and Oceanic Technology, Vol. 15, No. 2, 04.1998, p. 510-521.

Research output: Contribution to journalArticle

@article{ec292ace8d2f4df2a574744a0a9324f8,
title = "A spectral element solution of the shallow-water equations on multiprocessor computers",
abstract = "A shallow-water spectral element ocean model is implemented on multiple instruction multiple data, distributed memory parallel computers. A communications-minimizing partitioning algorithm for unstructured meshes, based on graph theory, is presented and is shown to improve the efficiency in a limited range of granularities. A domain decomposition implementation with an architecture-independent communications scheme, using message passing, is devised and tested on an nCUBE/2, a Cray T3D, and an IBM SP2. The implementation exhibits high efficiencies over a wide range of granularities. An order of magnitude analysis shows that, to leading order, the efficiency stays constant when KN2 grows proportionally to P, where K is the total number of elements, N is the order of the spectral truncation within an element, and P is the number of processors.",
author = "Curchitser, {Enrique N.} and Mohamed Iskandarani and Haidvogel, {Dale B.}",
year = "1998",
month = "4",
language = "English (US)",
volume = "15",
pages = "510--521",
journal = "Journal of Atmospheric and Oceanic Technology",
issn = "0739-0572",
publisher = "American Meteorological Society",
number = "2",

}

TY - JOUR

T1 - A spectral element solution of the shallow-water equations on multiprocessor computers

AU - Curchitser, Enrique N.

AU - Iskandarani, Mohamed

AU - Haidvogel, Dale B.

PY - 1998/4

Y1 - 1998/4

N2 - A shallow-water spectral element ocean model is implemented on multiple instruction multiple data, distributed memory parallel computers. A communications-minimizing partitioning algorithm for unstructured meshes, based on graph theory, is presented and is shown to improve the efficiency in a limited range of granularities. A domain decomposition implementation with an architecture-independent communications scheme, using message passing, is devised and tested on an nCUBE/2, a Cray T3D, and an IBM SP2. The implementation exhibits high efficiencies over a wide range of granularities. An order of magnitude analysis shows that, to leading order, the efficiency stays constant when KN2 grows proportionally to P, where K is the total number of elements, N is the order of the spectral truncation within an element, and P is the number of processors.

AB - A shallow-water spectral element ocean model is implemented on multiple instruction multiple data, distributed memory parallel computers. A communications-minimizing partitioning algorithm for unstructured meshes, based on graph theory, is presented and is shown to improve the efficiency in a limited range of granularities. A domain decomposition implementation with an architecture-independent communications scheme, using message passing, is devised and tested on an nCUBE/2, a Cray T3D, and an IBM SP2. The implementation exhibits high efficiencies over a wide range of granularities. An order of magnitude analysis shows that, to leading order, the efficiency stays constant when KN2 grows proportionally to P, where K is the total number of elements, N is the order of the spectral truncation within an element, and P is the number of processors.

UR - http://www.scopus.com/inward/record.url?scp=0000327482&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0000327482&partnerID=8YFLogxK

M3 - Article

VL - 15

SP - 510

EP - 521

JO - Journal of Atmospheric and Oceanic Technology

JF - Journal of Atmospheric and Oceanic Technology

SN - 0739-0572

IS - 2

ER -