Abstract
Artificial Neural Networks often rhyme with massive parallelism. We need parallelism to simulate neural nets and fully benefit of their promises but we do need an efficient way to implement it. So we propose a dedicated approach, general enough to encompass a large class of models. It consists in performing arithmetics oriented algorithms. That leads to the joint development of hardware and software tools: the SMART machine. SMART stands for Sparse Matrix Adaptive Recursive Transforms, the general framework we use to discribe ANN. In order to handle networks with a sparse or dynamic connections topology we propose an original architecture. Using high speed CMOS technology it will deliver up to 300 MFlops in a deck side format, supporting up to one million connections: from 1000 fully interconnected neuronlikes to 10 000 sparsely interconnected ones. With the user-friendly UNIX Workstation host and integrated high speed VME channels we get a Neurostation exhibiting a power to price ratio many times greater than a supercomputer one. In this paper, we discuss first the SMART approach and its appeal. Second, we shortly present an hardware architecture for SMART sparse calculus. Third, we address the programming issue and display early results. Finally we summarize the advantages of the proposed machine with performance forecastings.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
AMDHAL G., “The validity of the single processor approach to achieving large scale computing capabilities”, AFIPS conf proc., Vol. 30, 1967, SJCC, pp.483–485
BESSIERE P., CHAMS A., “Virtual model for ANN programming”, submited to INNC90, Paris (FRANCE), July 9–13, 1990
DERO T., ESCANDE P., MOULINOUX C., “A neuron-oriented programming environment”, NEURO-NIMES’89,Nîmes (FRANCE), November 13–16, 1989, EC2 Editeur, pp183–200
GEORGE A., LIU J., “Computer solutions of large sparse positive definite systems in computational mathematics”, Prentice-Hall, Englewood Cliffs. NJ, 1981
GROSSBERG S., “Non-linear neural networks: principles, mechanism, and architectures”, Neural Networks, Vol. 1, N°1, pp 17–61,1988
GUERIN A., HERAULT J., “An efficient vector and scalar processing structure for adaptive algorithms”, Signal Processing, IV, pp. 331–334, 1988
“Le traitement de l’information dans les structure nerveuses. Etude par la simulation numérique et éléctrique. Application au traitement des signaux”, thèse de doctoral ès sciences, I.N.P.Grenoble, 1980.
KOHONEN T., “Self organization and assiociative memory”, Springer, 1984
KRAFT T.T., FROSTROM A., MAC RITCHIE B., ROGERS A.E., “The specification of a concurrent back-propagation network architecture using actors”, NEURO-NIMES’89,Nîmes (FRANCE), November 13–16, 1989, EC2 Editeur, pp169–182
KUPERSTEIN M., “Neural model of adaptive hand-eye coordination for single postures”, Science, Vol. 239, 11 March 1988, pp 1308–1311
LALONDE W.R., GRAF D.H., “Neuroplanner: mechanisms for subcognitive control”, NEURONIMES’89, Nîmes (FRANCE), November 13–16, 1989, EC2 Editeur, pp 169–182
MELHEM R., “A systolic accelerator for the iterative solution os sparse linear systems”, IEEE Transaction on Computers,Vol. 38, N°11, November 1989, pp1591–1595
MUNTEAN T., “Supernode: architecture parallèle et dynamiquement reconfigurable de Transputers”, 11 èmes Journées francophones sur l’informatique,Nancy (FRANCE), janvier 1989, EC2 Editeur
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 1990 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Lawson, JC., Chams, A., Herault, J. (1990). Smart: How to Simulate Huge Networks. In: International Neural Network Conference. Springer, Dordrecht. https://doi.org/10.1007/978-94-009-0643-3_7
Download citation
DOI: https://doi.org/10.1007/978-94-009-0643-3_7
Publisher Name: Springer, Dordrecht
Print ISBN: 978-0-7923-0831-7
Online ISBN: 978-94-009-0643-3
eBook Packages: Springer Book Archive