An integrated performance visualizer for MPI/OpenMP programs

Jay Hoeflinger; Bob Kuhn; Wolfgang Nagel; Paul Petersen; Hrabri Rajic; Sanjiv Shah; Jeff Vetter; Michael Voss; Renee Woo

doi:10.1007/3-540-44587-0_5

An integrated performance visualizer for MPI/OpenMP programs

Jay Hoeflinger, Bob Kuhn, Wolfgang Nagel, Paul Petersen, Hrabri Rajic, Sanjiv Shah, Jeff Vetter, Michael Voss, Renee Woo

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

8 Scopus citations

Abstract

As cluster computing has grown, so has its use for large scientific calculations. Recently, many researchers have experimented with using MPI between nodes of a clustered machine and OpenMP within a node, to manage the use of parallel processing. Unfortunately, very few tools are available for doing an integrated analysis of an MPI/OpenMP program. KAI Software, Pallas GmbH and the US Department of Energy have partnered together to build such a tool, VGV. VGV is designed for doing scalable performance analysis - that is, to make the performance analysis process qualitatively the same for small cluster machines as it is for the largest ASCI systems. This paper describes VGV and gives a flavor of how to find performance problems using it.

Original language	English
Title of host publication	OpenMP Shared Memory Parallel Programming - International Workshop on OpenMP Applications and Tools,WOMPAT 2001, Proceedings
Editors	Rudolf Eigenmann, Michael J. Voss
Publisher	Springer Verlag
Pages	40-52
Number of pages	13
ISBN (Print)	9783540445876
DOIs	https://doi.org/10.1007/3-540-44587-0_5
State	Published - 2001
Externally published	Yes
Event	International Workshop on OpenMP Applications and Tools, WOMPAT 2001 - West Lafayette, United States Duration: Jul 30 2001 → Jul 31 2001

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	2104
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International Workshop on OpenMP Applications and Tools, WOMPAT 2001
Country/Territory	United States
City	West Lafayette
Period	07/30/01 → 07/31/01

Access to Document

10.1007/3-540-44587-0_5

Cite this

Hoeflinger, J., Kuhn, B., Nagel, W., Petersen, P., Rajic, H., Shah, S., Vetter, J., Voss, M., & Woo, R. (2001). An integrated performance visualizer for MPI/OpenMP programs. In R. Eigenmann, & M. J. Voss (Eds.), OpenMP Shared Memory Parallel Programming - International Workshop on OpenMP Applications and Tools,WOMPAT 2001, Proceedings (pp. 40-52). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2104). Springer Verlag. https://doi.org/10.1007/3-540-44587-0_5

Hoeflinger, Jay ; Kuhn, Bob ; Nagel, Wolfgang et al. / An integrated performance visualizer for MPI/OpenMP programs. OpenMP Shared Memory Parallel Programming - International Workshop on OpenMP Applications and Tools,WOMPAT 2001, Proceedings. editor / Rudolf Eigenmann ; Michael J. Voss. Springer Verlag, 2001. pp. 40-52 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{e6953ad3dfd74af29d18c3af4c410ecb,

title = "An integrated performance visualizer for MPI/OpenMP programs",

abstract = "As cluster computing has grown, so has its use for large scientific calculations. Recently, many researchers have experimented with using MPI between nodes of a clustered machine and OpenMP within a node, to manage the use of parallel processing. Unfortunately, very few tools are available for doing an integrated analysis of an MPI/OpenMP program. KAI Software, Pallas GmbH and the US Department of Energy have partnered together to build such a tool, VGV. VGV is designed for doing scalable performance analysis - that is, to make the performance analysis process qualitatively the same for small cluster machines as it is for the largest ASCI systems. This paper describes VGV and gives a flavor of how to find performance problems using it.",

author = "Jay Hoeflinger and Bob Kuhn and Wolfgang Nagel and Paul Petersen and Hrabri Rajic and Sanjiv Shah and Jeff Vetter and Michael Voss and Renee Woo",

note = "Publisher Copyright: {\textcopyright} Springer-Verlag Berlin Heidelberg 2001.; International Workshop on OpenMP Applications and Tools, WOMPAT 2001 ; Conference date: 30-07-2001 Through 31-07-2001",

year = "2001",

doi = "10.1007/3-540-44587-0_5",

language = "English",

isbn = "9783540445876",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "40--52",

editor = "Rudolf Eigenmann and Voss, {Michael J.}",

booktitle = "OpenMP Shared Memory Parallel Programming - International Workshop on OpenMP Applications and Tools,WOMPAT 2001, Proceedings",

}

Hoeflinger, J, Kuhn, B, Nagel, W, Petersen, P, Rajic, H, Shah, S, Vetter, J, Voss, M & Woo, R 2001, An integrated performance visualizer for MPI/OpenMP programs. in R Eigenmann & MJ Voss (eds), OpenMP Shared Memory Parallel Programming - International Workshop on OpenMP Applications and Tools,WOMPAT 2001, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 2104, Springer Verlag, pp. 40-52, International Workshop on OpenMP Applications and Tools, WOMPAT 2001, West Lafayette, United States, 07/30/01. https://doi.org/10.1007/3-540-44587-0_5

An integrated performance visualizer for MPI/OpenMP programs. / Hoeflinger, Jay; Kuhn, Bob; Nagel, Wolfgang et al.
OpenMP Shared Memory Parallel Programming - International Workshop on OpenMP Applications and Tools,WOMPAT 2001, Proceedings. ed. / Rudolf Eigenmann; Michael J. Voss. Springer Verlag, 2001. p. 40-52 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2104).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - An integrated performance visualizer for MPI/OpenMP programs

AU - Hoeflinger, Jay

AU - Kuhn, Bob

AU - Nagel, Wolfgang

AU - Petersen, Paul

AU - Rajic, Hrabri

AU - Shah, Sanjiv

AU - Vetter, Jeff

AU - Voss, Michael

AU - Woo, Renee

PY - 2001

Y1 - 2001

N2 - As cluster computing has grown, so has its use for large scientific calculations. Recently, many researchers have experimented with using MPI between nodes of a clustered machine and OpenMP within a node, to manage the use of parallel processing. Unfortunately, very few tools are available for doing an integrated analysis of an MPI/OpenMP program. KAI Software, Pallas GmbH and the US Department of Energy have partnered together to build such a tool, VGV. VGV is designed for doing scalable performance analysis - that is, to make the performance analysis process qualitatively the same for small cluster machines as it is for the largest ASCI systems. This paper describes VGV and gives a flavor of how to find performance problems using it.

AB - As cluster computing has grown, so has its use for large scientific calculations. Recently, many researchers have experimented with using MPI between nodes of a clustered machine and OpenMP within a node, to manage the use of parallel processing. Unfortunately, very few tools are available for doing an integrated analysis of an MPI/OpenMP program. KAI Software, Pallas GmbH and the US Department of Energy have partnered together to build such a tool, VGV. VGV is designed for doing scalable performance analysis - that is, to make the performance analysis process qualitatively the same for small cluster machines as it is for the largest ASCI systems. This paper describes VGV and gives a flavor of how to find performance problems using it.

UR - http://www.scopus.com/inward/record.url?scp=84974678406&partnerID=8YFLogxK

U2 - 10.1007/3-540-44587-0_5

DO - 10.1007/3-540-44587-0_5

M3 - Conference contribution

AN - SCOPUS:84974678406

SN - 9783540445876

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 40

EP - 52

BT - OpenMP Shared Memory Parallel Programming - International Workshop on OpenMP Applications and Tools,WOMPAT 2001, Proceedings

A2 - Eigenmann, Rudolf

A2 - Voss, Michael J.

PB - Springer Verlag

T2 - International Workshop on OpenMP Applications and Tools, WOMPAT 2001

Y2 - 30 July 2001 through 31 July 2001

ER -

Hoeflinger J, Kuhn B, Nagel W, Petersen P, Rajic H, Shah S et al. An integrated performance visualizer for MPI/OpenMP programs. In Eigenmann R, Voss MJ, editors, OpenMP Shared Memory Parallel Programming - International Workshop on OpenMP Applications and Tools,WOMPAT 2001, Proceedings. Springer Verlag. 2001. p. 40-52. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/3-540-44587-0_5

An integrated performance visualizer for MPI/OpenMP programs

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this