Paper Info
Title
The Influence of Operating Systems on the Performance of Collective Operations at Extreme Scale
Abstract
Abstract We investigate operating system noise, which we identify as one of the main reasons for a lack of synchronicity in par- allel 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 ar- tificially 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 in- terruption to the application, even if the probability of such an interruption is extremely small. We demonstrate that syn- chronizing the noise can significantly reduce its negative in- fluence. Keywords: microbenchmark, noise, petascale,
Year
DOI
Venue
2006
10.1109/CLUSTR.2006.311846
Barcelona
Keywords
DocType
ISSN
benchmark testing,noise measurement,operating systems (computers),parallel processing,software performance evaluation,extreme-scale platforms,microbenchmark,operating system noise,parallel applications,microbenchmark,noise,petascale,synchronicity
Conference
1552-5244 E-ISBN : 1-4244-0328-6
ISBN
Citations 
PageRank 
1-4244-0328-6
49
3.28
References 
Authors
11
4
Name
Order
Citations
PageRank
Pete Beckman182248.04
Kamil Iskra264246.46
Kazutomo Yoshii324918.53
Susan Coghlan429118.09