Klaus Obermayer, Helmut Heller, Helge Ritter, and Klaus Schulten.
Simulation of self-organizing neural nets: A comparision between a
transputer ring and a Connection Machine CM-2.
In Alan S. Wagner, editor, NATUG 3: Transputer Research and
Applications 3, pp. 95-106, Amsterdam, 1990. North American Transputer
Users Group, IOS Press.
OBER90B
In this contribution we describe parallel algorithms implementing Kohonen's "Self-organizing Feature Maps" on a Transputer systolic ring and on a Connection Machine CM-2. We implemented Kohonen's algorithm with the goal to study its ability for biological modelling and to explain the formation and the adaptive properties of topographic feature maps in the CNS of higher animals. Therefore our attention lies on networks with a high number of units receiving pattern vectors from a high dimensional input space. In the following we present benchmark studies (i) to measure the performance of the algorithm as a function of important network parameters on both systems, (ii) to compare the performance on the MIMD - Transputer systolic ring ("coarse-grained" implementation) with the performance on the SIMD - Connection Machine CM-2 ("fine-grained" implementation), and (iii) to measure the speedup of the algorithm depending on the number of physical processors used. The results are supplemented by measurements of communication times between the processors to allow predictions for related algorithms involving higher communication between units. A short overview over results obtained with both parallel machines is given.
Download Full Text
The manuscripts available on our site are provided for your personal
use only and may not be retransmitted or redistributed without written
permissions from the paper's publisher and author. You may not upload any
of this site's material to any public server, on-line service, network, or
bulletin board without prior written permission from the publisher and
author. You may not make copies for any commercial purpose. Reproduction
or storage of materials retrieved from this web site is subject to the
U.S. Copyright Act of 1976, Title 17 U.S.C.