loading...
 This Article 
   
 Share 
   
 Bibliographic References 
   
 Add to: 
 
Digg
Furl
Spurl
Blink
Simpy
Google
Del.icio.us
Y!MyWeb
 
 Search 
   
2006 IEEE International Conference on Cluster Computing
The Influence of Operating Systems on the Performance of Collective Operations at Extreme Scale
Barcelona
September 25-September 28
ISBN: 1-4244-0327-8
ASCII Text
x 
 
P. Beckman, K. Iskra, K. Yoshii, S. Coghlan, "The Influence of Operating Systems on the Performance of Collective Operations at Extreme Scale," Cluster Computing, IEEE International Conference on, pp. 1-12, 2006 IEEE International Conference on Cluster Computing, 2006.
 
 
BibTex
x
 
@article{ 10.1109/CLUSTR.2006.311846,
author = {P. Beckman and K. Iskra and K. Yoshii and S. Coghlan},
title = {The Influence of Operating Systems on the Performance of Collective Operations at Extreme Scale},
journal ={Cluster Computing, IEEE International Conference on},
volume = {0},
year = {2006},
issn = {1552-5244},
pages = {1-12},
doi = {http://doi.ieeecomputersociety.org/10.1109/CLUSTR.2006.311846},
publisher = {IEEE Computer Society},
address = {Los Alamitos, CA, USA},
}
 
 
RefWorks Procite/RefMan/Endnote
x 
 
TY - CONF
JO - Cluster Computing, IEEE International Conference on
TI - The Influence of Operating Systems on the Performance of Collective Operations at Extreme Scale
SN - 1552-5244
SP1
EP12
A1 - P. Beckman,
A1 - K. Iskra,
A1 - K. Yoshii,
A1 - S. Coghlan,
PY - 2006
KW - extreme-scale platforms
KW - operating system noise
KW - parallel applications
KW - microbenchmark
VL - 0
JA - Cluster Computing, IEEE International Conference on
ER -
 
P. Beckman, Argonne Nat. Lab., Math.&Comput. Sci. Div., Argonne, IL
K. Iskra, Argonne Nat. Lab., Math.&Comput. Sci. Div., Argonne, IL
K. Yoshii, Argonne Nat. Lab., Math.&Comput. Sci. Div., Argonne, IL
S. Coghlan, Argonne Nat. Lab., Math.&Comput. Sci. Div., Argonne, IL
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
Index Terms:
extreme-scale platforms, operating system noise, parallel applications, microbenchmark
Citation:
P. Beckman, K. Iskra, K. Yoshii, S. Coghlan, "The Influence of Operating Systems on the Performance of Collective Operations at Extreme Scale," cluster, pp.1-12, 2006 IEEE International Conference on Cluster Computing, 2006
Usage of this product signifies your acceptance of the Terms of Use.