Evaluation of the two different interconnection networks of the CNAPS neurocomputer

  • Bertrand Granado
  • Patrick Garda
Poster Presentations 2 Implementations
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1112)


In this paper, we presented a general mapping algorithm for MLP onto the CNAPS neurocomputer. We derived an analytical model of the performances of CNAPS for the two schemes, local and full, of interconnections between layers. We compared the two interconnection networks of CNAPS for their implementation. Finally we gave the performance of two neural systems for character recognition. It should be observed that the performances achieved for the simulation of LeNet are only 2 to 4 times slower than those achieved by dedicated hardware such as ANNA.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Jim Baley and Dan Hammerstrom. Why vlsi implementation of associative vlcns require connection multiplexing. In Proceedings of the IEEE 2nd Annual International Conference on Neural Network, pages 112–119, 1988.Google Scholar
  2. 2.
    Y. Le Cun, B. Boser, J.S. Denker, D.henderson, R.E. Howard, W. hubbard, and L.J. Jackel. Handwritten digit recognition with a back-propagation network. In Neural Information Process and System, pages 396–404, 1990.Google Scholar
  3. 3.
    Antoine Dupuy and Patrick Garda. Répartition optimale de réseaux de neurones en temps linéaire sur une machine parallèle communicant par bus sécable. In Actes de renpar'5 — 5ieme rencontres Francophones du Parallélisme, Mai 1993.Google Scholar
  4. 4.
    Bertrand Granado and Patrick Garda. Une méthode de prédiction des temps de simulation des perceptrons multi-couches sur des machines à parallélisme de données. In Actes de Renpar'7, Juin 1995.Google Scholar
  5. 5.
    Matthew Griffin, Gary Tahara, Kurt Knorpp, Ray Pinkham, Bob Riley, Dan Hamerstrom, and Eric Means. An 11 million transistor digital neural network execution engine. In Proceedings of IEEE International Solid-State Circuits Conference, 1991.Google Scholar
  6. 6.
    R.W Hockney and C.R Jesshope. Parallel Computer 2. Adam Hilger, Bristol and Philadelphia, 1988.Google Scholar
  7. 7.
    Jacques-Olivier Klein, Hubert Pujol, and Patrick Garda. Simulation de la machine de boltzmann en temps réel. Traitement du Signal, Juillet 1994.Google Scholar
  8. 8.
    Dean Mueller and Dan Hammerstrom. A neural network systems component. In Proceedings od International Conference on Neural Network, pages 1258–1264, 1993.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1996

Authors and Affiliations

  • Bertrand Granado
    • 1
  • Patrick Garda
    • 1
  1. 1.Laboratoire d'Electronique Analogique et Micro-ondes Boîte 203Université Pierre et Marie CurieParis Cedex 05

Personalised recommendations