By Topic

Statistical analysis of message passing programs to guide computer design

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$33 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

2 Author(s)
W. E. Cohen ; Dept. of Electr. & Comput. Eng., Alabama Univ., Huntsville, AL, USA ; B. Ali Mahafzah

Little data exists on how message passing programs use parallel computers. The behavior of these programs can strongly influence design decisions made for future computer systems. The computer designer's use of incorrect assumptions about program behavior can degrade performance. In many cases simple statistical parameters describing characteristics such as message sizes, destinations, sources, and times between sends would give the designers of the communication libraries and the computer hardware great insight into how the hardware is used by actual programs. Techniques of collecting statistical information about the communication characteristics for system design have been applied to the parallel version of the NAS benchmarks. The paper describes the statistical data collected for multiprocessor runs of the NAS 2.1 benchmarks and some of the characteristics observed in that data

Published in:

System Sciences, 1998., Proceedings of the Thirty-First Hawaii International Conference on  (Volume:7 )

Date of Conference:

6-9 Jan 1998