Paper published in a journal (Scientific congresses, symposiums and conference proceedings)
Deskilling HPL - Using an Evolutionary Algorithm to Automate Cluster Benchmarking
Dunlop, Dominic; Varrette, Sébastien; Bouvry, Pascal
2009In Proceedings of 8th Intl. conf. on Parallel Processing and Applied Mathematics (PPAM 2009), 6068 (LNCS), p. 102--114
Peer reviewed
 

Files


Full Text
2009_PPAM_DeskillingHPL.pdf
Publisher postprint (325.33 kB)
Download

All documents in ORBilu are protected by a user license.

Send to



Details



Keywords :
HPL; Performance Evaluation; Evolutionary Algorithm
Abstract :
[en] The High-Performance Linpack (HPL) benchmark is the ac- cepted standard for measuring the capacity of the world’s most powerful computers, which are ranked twice yearly in the Top 500 List. Since just a small deficit in performance can cost a computer several places, it is impor- tant to tune the benchmark to obtain the best possible result. However, the adjustment of HPL’s seventeen configuration parameters to obtain maximum performance is a time-consuming task that must be performed by hand. In a previous paper, we provided a preliminary study that proposed the tuning of HPL parameters by means of an Evolutionary Algorithm. The approach was validated on a small cluster hosted at the University of Luxembourg. In this article, we extend this initial work by describing Acbea, a fully-automatic benchmark tuning tool that performs both the configuration and installation of HPL followed by an automatic search for optimized parameters that will lead to the best benchmark results. Experiments have been conducted to validate this tool on several clusters, exploiting in particular the Grid’5000 infrastructure.
Research center :
ULHPC - University of Luxembourg: High Performance Computing
Disciplines :
Computer science
Identifiers :
UNILU:UL-CONFERENCE-2009-807
Author, co-author :
Dunlop, Dominic ;  University of Luxembourg > Faculty of Science, Technology and Communication (FSTC) > Computer Science and Communications Research Unit (CSC)
Varrette, Sébastien ;  University of Luxembourg > Faculty of Science, Technology and Communication (FSTC) > Computer Science and Communications Research Unit (CSC)
Bouvry, Pascal ;  University of Luxembourg > Faculty of Science, Technology and Communication (FSTC) > Computer Science and Communications Research Unit (CSC)
Language :
English
Title :
Deskilling HPL - Using an Evolutionary Algorithm to Automate Cluster Benchmarking
Publication date :
September 2009
Event name :
8th Intl. Conf. on Parallel Processing and Applied Mathematics (PPAM 2009)
Event place :
Wroclaw, Poland
Event date :
from 13-09-2009 to 16-09-2009
Audience :
International
Journal title :
Proceedings of 8th Intl. conf. on Parallel Processing and Applied Mathematics (PPAM 2009)
Publisher :
Springer Verlag
Volume :
6068
Issue :
LNCS
Pages :
102--114
Peer reviewed :
Peer reviewed
Additional URL :
Available on ORBilu :
since 02 December 2013

Statistics


Number of views
110 (3 by Unilu)
Number of downloads
106 (2 by Unilu)

WoS citations
 
1

Bibliography


Similar publications



Contact ORBilu