skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: The influence of operating systems on the performance of collective operations at extreme scale.

Abstract

We investigate operating system noise, which we identify as one of the main reasons for a lack of synchronicity in parallel applications. Using a microbenchmark, we measure the noise on several contemporary platforms and find that, even with a general-purpose operating system, noise can be limited if certain precautions are taken. We then inject artificially generated noise into a massively parallel system and measure its influence on the performance of collective operations. Our experiments indicate that on extreme-scale platforms, the performance is correlated with the largest interruption to the application, even if the probability of such an interruption is extremely small. We demonstrate that synchronizing the noise can significantly reduce its negative influence.

Authors:
; ; ; ;
Publication Date:
Research Org.:
Argonne National Lab. (ANL), Argonne, IL (United States)
Sponsoring Org.:
USDOE Office of Science (SC)
OSTI Identifier:
982596
Report Number(s):
ANL/MCS/CP-118581
TRN: US201015%%1206
DOE Contract Number:  
DE-AC02-06CH11357
Resource Type:
Conference
Resource Relation:
Conference: 2006 IEEE International Conference on Cluster Computing; Sep. 25, 2006 - Sep. 28, 2006; Barcelona, Spain
Country of Publication:
United States
Language:
ENGLISH
Subject:
97 MATHEMATICAL METHODS AND COMPUTING; 99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; MEETINGS; NOISE; PERFORMANCE; PROBABILITY; USES

Citation Formats

Beckman, P., Iskra, K., Yoshii, K., Coghlan, S., and Mathematics and Computer Science. The influence of operating systems on the performance of collective operations at extreme scale.. United States: N. p., 2006. Web.
Beckman, P., Iskra, K., Yoshii, K., Coghlan, S., & Mathematics and Computer Science. The influence of operating systems on the performance of collective operations at extreme scale.. United States.
Beckman, P., Iskra, K., Yoshii, K., Coghlan, S., and Mathematics and Computer Science. Sun . "The influence of operating systems on the performance of collective operations at extreme scale.". United States. doi:.
@article{osti_982596,
title = {The influence of operating systems on the performance of collective operations at extreme scale.},
author = {Beckman, P. and Iskra, K. and Yoshii, K. and Coghlan, S. and Mathematics and Computer Science},
abstractNote = {We investigate operating system noise, which we identify as one of the main reasons for a lack of synchronicity in parallel applications. Using a microbenchmark, we measure the noise on several contemporary platforms and find that, even with a general-purpose operating system, noise can be limited if certain precautions are taken. We then inject artificially generated noise into a massively parallel system and measure its influence on the performance of collective operations. Our experiments indicate that on extreme-scale platforms, the performance is correlated with the largest interruption to the application, even if the probability of such an interruption is extremely small. We demonstrate that synchronizing the noise can significantly reduce its negative influence.},
doi = {},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Sun Jan 01 00:00:00 EST 2006},
month = {Sun Jan 01 00:00:00 EST 2006}
}

Conference:
Other availability
Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.

Save / Share: