Interpretive performance prediction for High Performance application development

Manish Parashar, Salim A Hariri

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Software development for high Performance (parallel/distributed) Computing (HPC) is a non-trivial process; its complexity can be primarily attributed to the increased degrees of freedom that have to be resolved and tuned in such an environment. Performance prediction tools enable a developer to evaluate various available design alternatives and can assist in HPC application software development. In this paper we first present a novel `interpretive' approach for accurate and cost-effective performance prediction. The approach has been used to develop an interpretive HPF/Fortran 90D application performance prediction framework. The accuracy and usability of the performance prediction framework are experimentally validated. We then outline the stages typically encountered during application software development for parallel/distributed HPC and highlight the significance and requirements of a performance prediction tool at the relevant stages. Numerical results using benchmarking kernels and application codes are presented to demonstrate the application of the interpretive performance prediction framework at different stages of the software development process.

Original languageEnglish (US)
Title of host publicationProceedings of the Hawaii International Conference on System Sciences
Editors Anon
PublisherIEEE Comp Soc
Pages462-471
Number of pages10
Volume1
StatePublished - 1997
Externally publishedYes
EventProceedings of the 1997 30th Annual Hawaii International Conference on System Sciences. Part 1 (of 6) - Wailea, HI, USA
Duration: Jan 7 1997Jan 10 1997

Other

OtherProceedings of the 1997 30th Annual Hawaii International Conference on System Sciences. Part 1 (of 6)
CityWailea, HI, USA
Period1/7/971/10/97

Fingerprint

Software engineering
Application programs
Distributed computer systems
Benchmarking
Costs

ASJC Scopus subject areas

  • Software
  • Industrial and Manufacturing Engineering

Cite this

Parashar, M., & Hariri, S. A. (1997). Interpretive performance prediction for High Performance application development. In Anon (Ed.), Proceedings of the Hawaii International Conference on System Sciences (Vol. 1, pp. 462-471). IEEE Comp Soc.

Interpretive performance prediction for High Performance application development. / Parashar, Manish; Hariri, Salim A.

Proceedings of the Hawaii International Conference on System Sciences. ed. / Anon. Vol. 1 IEEE Comp Soc, 1997. p. 462-471.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Parashar, M & Hariri, SA 1997, Interpretive performance prediction for High Performance application development. in Anon (ed.), Proceedings of the Hawaii International Conference on System Sciences. vol. 1, IEEE Comp Soc, pp. 462-471, Proceedings of the 1997 30th Annual Hawaii International Conference on System Sciences. Part 1 (of 6), Wailea, HI, USA, 1/7/97.
Parashar M, Hariri SA. Interpretive performance prediction for High Performance application development. In Anon, editor, Proceedings of the Hawaii International Conference on System Sciences. Vol. 1. IEEE Comp Soc. 1997. p. 462-471
Parashar, Manish ; Hariri, Salim A. / Interpretive performance prediction for High Performance application development. Proceedings of the Hawaii International Conference on System Sciences. editor / Anon. Vol. 1 IEEE Comp Soc, 1997. pp. 462-471
@inproceedings{93d0dbd5e0804722804e9ad2b4bf11ea,
title = "Interpretive performance prediction for High Performance application development",
abstract = "Software development for high Performance (parallel/distributed) Computing (HPC) is a non-trivial process; its complexity can be primarily attributed to the increased degrees of freedom that have to be resolved and tuned in such an environment. Performance prediction tools enable a developer to evaluate various available design alternatives and can assist in HPC application software development. In this paper we first present a novel `interpretive' approach for accurate and cost-effective performance prediction. The approach has been used to develop an interpretive HPF/Fortran 90D application performance prediction framework. The accuracy and usability of the performance prediction framework are experimentally validated. We then outline the stages typically encountered during application software development for parallel/distributed HPC and highlight the significance and requirements of a performance prediction tool at the relevant stages. Numerical results using benchmarking kernels and application codes are presented to demonstrate the application of the interpretive performance prediction framework at different stages of the software development process.",
author = "Manish Parashar and Hariri, {Salim A}",
year = "1997",
language = "English (US)",
volume = "1",
pages = "462--471",
editor = "Anon",
booktitle = "Proceedings of the Hawaii International Conference on System Sciences",
publisher = "IEEE Comp Soc",

}

TY - GEN

T1 - Interpretive performance prediction for High Performance application development

AU - Parashar, Manish

AU - Hariri, Salim A

PY - 1997

Y1 - 1997

N2 - Software development for high Performance (parallel/distributed) Computing (HPC) is a non-trivial process; its complexity can be primarily attributed to the increased degrees of freedom that have to be resolved and tuned in such an environment. Performance prediction tools enable a developer to evaluate various available design alternatives and can assist in HPC application software development. In this paper we first present a novel `interpretive' approach for accurate and cost-effective performance prediction. The approach has been used to develop an interpretive HPF/Fortran 90D application performance prediction framework. The accuracy and usability of the performance prediction framework are experimentally validated. We then outline the stages typically encountered during application software development for parallel/distributed HPC and highlight the significance and requirements of a performance prediction tool at the relevant stages. Numerical results using benchmarking kernels and application codes are presented to demonstrate the application of the interpretive performance prediction framework at different stages of the software development process.

AB - Software development for high Performance (parallel/distributed) Computing (HPC) is a non-trivial process; its complexity can be primarily attributed to the increased degrees of freedom that have to be resolved and tuned in such an environment. Performance prediction tools enable a developer to evaluate various available design alternatives and can assist in HPC application software development. In this paper we first present a novel `interpretive' approach for accurate and cost-effective performance prediction. The approach has been used to develop an interpretive HPF/Fortran 90D application performance prediction framework. The accuracy and usability of the performance prediction framework are experimentally validated. We then outline the stages typically encountered during application software development for parallel/distributed HPC and highlight the significance and requirements of a performance prediction tool at the relevant stages. Numerical results using benchmarking kernels and application codes are presented to demonstrate the application of the interpretive performance prediction framework at different stages of the software development process.

UR - http://www.scopus.com/inward/record.url?scp=0031348029&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031348029&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0031348029

VL - 1

SP - 462

EP - 471

BT - Proceedings of the Hawaii International Conference on System Sciences

A2 - Anon, null

PB - IEEE Comp Soc

ER -