• Documents
  • Authors
  • Tables
  • Log in
  • Sign up
  • MetaCart
  • DMCA
  • Donate

CiteSeerX logo

Advanced Search Include Citations
Advanced Search Include Citations | Disambiguate

Ultra-scale parallel discrete event applications. Oak Ridge National Laboratory. Available from: http://www.bgconsortium.org/ [accessed 16 March 2008] Perumalla, K.S., 2007. Scaling Time Warp-based discrete event execution to 10**4 processors on a blue ge (2006)

by K S Perumalla
Venue:Proceedings of the 4th international conference on Computing frontiers, pp 69–76
Add To MetaCart

Tools

Sorted by:
Results 1 - 1 of 1

Analysis of time warp on a 32,768 processor ibm blue gene/l supercomputer

by Akintayo O. Holder, Christopher D. Carothers - in 2008 Proceedings European Modeling and Simulation Symposium (EMSS , 2008
"... The aim of our work is to investigate the performance and overall scalability of an optimistic discrete-event simulator on a Blue Gene/L supercomputer. We find that strong scaling out to 16,384 processors is possible. In terms of event-rate, we observed 853 million events per second on 16,384 proces ..."
Abstract - Cited by 7 (4 self) - Add to MetaCart
The aim of our work is to investigate the performance and overall scalability of an optimistic discrete-event simulator on a Blue Gene/L supercomputer. We find that strong scaling out to 16,384 processors is possible. In terms of event-rate, we observed 853 million events per second on 16,384 processors for the PHOLD benchmark. This is 1.5 times faster than any previously reported PDES synchronization protocol for PHOLD executing on a Blue Gen/L supercomputer (e.g. conservative, optimistic or hybrid). Additionally, we observed 2.47 billion events per second for a PCS telephone network model when executed on 32,768 processors. To the best of our knowledge, this is the first multi-billion event rate achieved for any Time Warp model executing on a Blue Gene/L supercomputer.
(Show Context)

Citation Context

...l case to be areasonable situation in a model. 4 EXPERIMENTAL RESULTS PHOLD is a synthetic benchmark, commonly used for testing the performance of Time Warp simulators (Chen and Szymanski 2005) and (=-=Perumalla 2006-=-). PHOLD has minimal event processing, minimal look ahead due to event scheduling being based on a random distribution and a random communication pattern. PHOLD can be configured by changing the event...

Powered by: Apache Solr
  • About CiteSeerX
  • Submit and Index Documents
  • Privacy Policy
  • Help
  • Data
  • Source
  • Contact Us

Developed at and hosted by The College of Information Sciences and Technology

© 2007-2019 The Pennsylvania State University