By Topic

Efficient parallel prefix algorithms on fully connected message-passing computers

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
$31 $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)
Yen-Chun Lin ; Dept. of Electron. Eng., Nat. Taiwan Inst. of Technol., Taipei, Taiwan ; Lin, C.M.

Given n values ν(0), ν(1), …, ν(n-1) and an associative binary operation, denoted by o, the prefix problem is to compute the n prefixes ν(0) o ν(1) o…0 ν(i), 0⩽i⩽n-1. We are interested in prefix computation on message-passing fully connected multicomputers, in which each processor can only send or receive a message to or from any other processor in a communication step, in as few communication steps as possible. An algorithm is presented to solve the prefix problem on a system of n processors in no more than [1.44 log2 n]+1 communication steps. Then, to explore the possibility of obtaining a faster algorithm, a class of algorithms is presented, It is shown that the algorithm in this class requiring the fewest communication steps is equivalent to the first algorithm presented. An algorithm using p<n processors is also presented; it is time-optimal and cost-optimal when p=Θ(n/log n)

Published in:

High Performance Computing, 1996. Proceedings. 3rd International Conference on

Date of Conference:

19-22 Dec 1996