Distributed processing of a regional prediction model

K. W. Johnson; J. Bauer; G. A. Riccardi; K. K. Droegemeier; Ming  Xue

doi:10.1175/1520-0493(1994)122<2558:DPOARP>2.0.CO;2

Distributed processing of a regional prediction model

K. W. Johnson, J. Bauer, G. A. Riccardi, K. K. Droegemeier, Ming Xue

Research output: Contribution to journal › Article › peer-review

Abstract

A domain decomposition strategy using a network of workstations produced a significant decrease in elapsed time and increase in problem size relative to a single-workstation run. The peformance of the resulting program is described by derived formulas (collectively known as a performance model), which predict the execution time and speedup for different numbers of processors and problem sizes. The interprocessor communication speeds are shown to be the major obstacle to achieving full processor use. The effect of faster communication networks on parallel performance is predicted based on this performance model. Parallelization experiments using the ARPS code were run on a cluster of IBM RS6000 workstations connected via Ethernet. -from Authors

Original language	English (US)
Pages (from-to)	2558-2572
Number of pages	15
Journal	Monthly Weather Review
Volume	122
Issue number	11
DOIs	https://doi.org/10.1175/1520-0493(1994)122<2558:DPOARP>2.0.CO;2
State	Published - 1994
Externally published	Yes

ASJC Scopus subject areas

Atmospheric Science

Online availability

10.1175/1520-0493(1994)122<2558:DPOARP>2.0.CO;2

Library availability

Discover UIUC Full Text

Cite this

@article{7b6a1653ed7d440aa310342b83c2dd07,

title = "Distributed processing of a regional prediction model",

abstract = "A domain decomposition strategy using a network of workstations produced a significant decrease in elapsed time and increase in problem size relative to a single-workstation run. The peformance of the resulting program is described by derived formulas (collectively known as a performance model), which predict the execution time and speedup for different numbers of processors and problem sizes. The interprocessor communication speeds are shown to be the major obstacle to achieving full processor use. The effect of faster communication networks on parallel performance is predicted based on this performance model. Parallelization experiments using the ARPS code were run on a cluster of IBM RS6000 workstations connected via Ethernet. -from Authors",

author = "Johnson, {K. W.} and J. Bauer and Riccardi, {G. A.} and Droegemeier, {K. K.} and Ming Xue",

note = "Funding Information: Acknowledgements We thank our patients and their families for participating in this study and the members of Taylor laboratory and Marie-Jos{\'e}e and Henry R. Kravis Center for Molecular Oncology for discussions and support. We thank the Jonathan and Mindy Gray Foundation for their support from project inception. This work was also supported by National Institutes of Health (NIH) awards P30 CA008748, U54 OD020355 (D.B.S., B.S.T.), R01 CA207244 (D.M.H., B.S.T.), P50 CA092629 (D.B.S., B.S.T.), R01 CA204749 (B.S.T.); and the American Cancer Society (RSG-15-067-01-TBG), Cycle for Survival, Sontag Foundation, Prostate Cancer Foundation, Anna Fuller Fund, and the Josie Robertson Foundation (B.S.T.).",

year = "1994",

doi = "10.1175/1520-0493(1994)122<2558:DPOARP>2.0.CO;2",

language = "English (US)",

volume = "122",

pages = "2558--2572",

journal = "Monthly Weather Review",

issn = "0027-0644",

publisher = "American Meteorological Society",

number = "11",

}

TY - JOUR

T1 - Distributed processing of a regional prediction model

AU - Johnson, K. W.

AU - Bauer, J.

AU - Riccardi, G. A.

AU - Droegemeier, K. K.

AU - Xue, Ming

N1 - Funding Information: Acknowledgements We thank our patients and their families for participating in this study and the members of Taylor laboratory and Marie-Josée and Henry R. Kravis Center for Molecular Oncology for discussions and support. We thank the Jonathan and Mindy Gray Foundation for their support from project inception. This work was also supported by National Institutes of Health (NIH) awards P30 CA008748, U54 OD020355 (D.B.S., B.S.T.), R01 CA207244 (D.M.H., B.S.T.), P50 CA092629 (D.B.S., B.S.T.), R01 CA204749 (B.S.T.); and the American Cancer Society (RSG-15-067-01-TBG), Cycle for Survival, Sontag Foundation, Prostate Cancer Foundation, Anna Fuller Fund, and the Josie Robertson Foundation (B.S.T.).

PY - 1994

Y1 - 1994

N2 - A domain decomposition strategy using a network of workstations produced a significant decrease in elapsed time and increase in problem size relative to a single-workstation run. The peformance of the resulting program is described by derived formulas (collectively known as a performance model), which predict the execution time and speedup for different numbers of processors and problem sizes. The interprocessor communication speeds are shown to be the major obstacle to achieving full processor use. The effect of faster communication networks on parallel performance is predicted based on this performance model. Parallelization experiments using the ARPS code were run on a cluster of IBM RS6000 workstations connected via Ethernet. -from Authors

AB - A domain decomposition strategy using a network of workstations produced a significant decrease in elapsed time and increase in problem size relative to a single-workstation run. The peformance of the resulting program is described by derived formulas (collectively known as a performance model), which predict the execution time and speedup for different numbers of processors and problem sizes. The interprocessor communication speeds are shown to be the major obstacle to achieving full processor use. The effect of faster communication networks on parallel performance is predicted based on this performance model. Parallelization experiments using the ARPS code were run on a cluster of IBM RS6000 workstations connected via Ethernet. -from Authors

UR - http://www.scopus.com/inward/record.url?scp=0028558080&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0028558080&partnerID=8YFLogxK

U2 - 10.1175/1520-0493(1994)122<2558:DPOARP>2.0.CO;2

DO - 10.1175/1520-0493(1994)122<2558:DPOARP>2.0.CO;2

M3 - Article

AN - SCOPUS:0028558080

SN - 0027-0644

VL - 122

SP - 2558

EP - 2572

JO - Monthly Weather Review

JF - Monthly Weather Review

IS - 11

ER -

Distributed processing of a regional prediction model

Abstract

ASJC Scopus subject areas

Online availability

Library availability

Related links

Fingerprint

Cite this