The Community for Technology Leaders
RSS Icon
Subscribe
Issue No.08 - August (1998 vol.9)
pp: 758-768
ABSTRACT
<p><b>Abstract</b>—In <it>s</it>-to-<it>p</it> broadcasting, <it>s</it> processors in a <it>p</it>-processor machine contain a message to be broadcast to all the processors, 1 ≤<it>s</it>≤<it>p</it>. We present a number of different broadcasting algorithms that handle all ranges of <it>s</it>. We show how the performance of each algorithm is influenced by the distribution of the <it>s</it> source processors and by the relationships between the distribution and the characteristics of the interconnection network. For the Intel Paragon we show that for each algorithm and machine dimension there exist ideal distributions and distributions on which the performance degrades. For the Cray T3D we also demonstrate dependencies between distributions and machine sizes. To reduce the dependence of the performance on the distribution of sources, we propose a repositioning approach. In this approach, the initial distribution is turned into an ideal distribution of the target broadcasting algorithm. We report experimental results for the Intel Paragon and Cray T3D and discuss scalability and performance.</p>
INDEX TERMS
Broadcasting, communication operations, message-passing MPPs, scalability.
CITATION
Susanne E. Hambrusch, Ashfaq A. Khokhar, Yi Liu, "Scalable S-To-P Broadcasting on Message-Passing MPPs", IEEE Transactions on Parallel & Distributed Systems, vol.9, no. 8, pp. 758-768, August 1998, doi:10.1109/71.706048
20 ms
(Ver 2.0)

Marketing Automation Platform Marketing Automation Tool