Abstract
The Edinburgh Concurrent Supercomputer Project is built around a Meiko Computing Surface, with presently some 400 floating point transputers and 1.6 Gbytes of memory. The first part of this paper gives a brief overview of the Project’s origins and status. In the second part we review work in neural network models, including analogue neurons for image restoration, studies of texture discrimination and protein structure predictions using a multi-layer perceptron simulator. The problem of optimization of machine topology is also discussed in the context of irregular graphs and genetic algorithms.
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
Bowler, K.C., Bruce, A.D., Kenway, R.D., Pawley, G.S., Wallace, DJ. and McKendrick, A., Scientific Computation on the Edinburgh DAPs, University of Edinburgh Report, December 1987.
Bowler, K.C., Kénway, R.D., Pawley, G.S. and Roweth, D., An Introduction to Occam 2 Programming, Chartwell-Bratt, Bromley 1987.
Wexler, J. and Wilson, G.V., Edinburgh Concurrent Supercomputer Project Directory, 1988.
Norman, M.G. and Fisher, R.B., Surface tracking within three dimensional datasets, using a generalised message passing harness, in Developments using occam (J. Kerridge, ed.) pp77–82, IOS Amsterdam, 1988.
Norman, M.G. and Wilson S., The TITCH User Guide; Clarke, L.J., The Tiny User Guide, available from the ECS Project, Edinburgh University Computing Service.
Prior, D., Radcliffe, N.J., Norman, M.G. and Clarke, L. J., Concurrency: Practice and Experience, to appear.
Valiant, L.G., Optimally universal parallel computers, in Scientific Applications of Multiprocessors (R.J. Elliott and C.A.R. Hoare, eds.) pp. 17–20, Prentice Hall International Series in Computer Science 1989.
Clarke, L. J., Rian User Guide, available from ECS Project, Edinburgh University Computing Service.
Norman, M.G. and Maclachlan, S., submitted to Conf. on Engineering Applications of Transputers, Liverpool, 1989.
Hopfield, J.J. and Tank, D., Neural computation on decisions in optimisation problems, Biol. Cyber. 52, 141–152 (1984).
Wilson, G.V. and Pawley, G.S., On the stability of the travelling salesman problem of Hopfield and Tank, Biol. Cyber. 58 63–70 (1988).
Tank, D.W. and Hopfield, J. J., AT&T Bell Labs preprint (1985).
Fox, G.C. and Furmanski, W., The physical structure of concurrent problems and concurrent computers, in Scientific Applications of Multiprocessors (RJ. Elliott and C.A.R. Hoare, eds.) pp. 55–88, Prentice Hall International Series in Computer Science 1989.
Geman, S. and Geman, D., Stochastic relaxation, Gibbs distributions and the Bayesian restoration of images, IEEE Trans PAMI 5, 721–741 (1984).
Murray, D.W., Kashko, A. and Buxton, B., A parallel approach to the picture restoration algorithm of Geman and Geman, IVC 3 133–142 (1985).
Kirkpatrick, S., Gellat, C.D. and Vecchi, M.P., Optimisation by simulated annealing, Science 220, 671–680 (1983).
Forrest, B.M., Restoration of binary images using networks of analogue neurons, in Parallel Architectures and Computer Vision (I. Page, ed.) pp. 19–31, Oxford University Press 1988.
Simmen, M. and Wilson G.V., A comparison of two parallel implementations of the Durbin and Willshaw algorithm for solving the travelling salesman problem, submitted to Concurrency: Practice and Experience.
Richards, G.D., Implementation of back-propagation on a transputer array, in Proc. 8th Technical Meeting of the Occam User Group (J. Kerridge, ed.) pp. 173–179, IOS Amsterdam 1988.
Richards, G.D., Documentation for Rhwydwaith, available from the ECS project, Edinburgh University Computing Service.
Smieja, F.J. and Richards, G.D., Hard learning the easy way — backpropagation with deformation, Complex Systems, to appear.
Dodd, N., Texture discrimination using multi-layer perceptrons, Pattern Recognition Letters, in press.
Brodatz, P., Textures — A Photographic Album for Artists and Designers, Dover, New York (1966).
Qian, N. and Sejnowski, T., Predicting the secondary structure of globular proteins using neural network models, J. Mol. Biol. 2020, 865–864 (1988).
Smieja, FJ., The significance of underlying correlations in the training of a layered net, presented at the INNS Conf., Boston, 1988. Edinburgh preprint, unpublished.
Smieja, FJ., MLP solutions, generalisation, and hidden unit representations, in Proc. DANIP Workshop, Bonn, 1989
Muhlenbein, H., Gorges-Schleuter, M. and Kramer, O., Evolution algorithms in combinatorial optimisation, Parallel Computing 7 65–85 (1985).
Radcliffe, N J., Early clustering around optima, Edinburgh preprint in preparation.
Norman, M.G., A genetic approach to topology optimisation for multiprocessor architectures, submitted to Parallel Computing
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1990 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Norman, M.G. et al. (1990). Neural Network Applications in the Edinburgh Concurrent Supercomputer Project. In: Soulié, F.F., Hérault, J. (eds) Neurocomputing. NATO ASI Series, vol 68. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-76153-9_24
Download citation
DOI: https://doi.org/10.1007/978-3-642-76153-9_24
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-76155-3
Online ISBN: 978-3-642-76153-9
eBook Packages: Springer Book Archive