Abstract
Neurons of the cortical tissue in a mammalian brain are connected in an extremely sparse and random fashion. The paper presents efficient methods for a parallel simulation of neural networks modeling this connection scheme on a CNAPS SIMD neurocomputer. Appropriate algorithms and data structures are introduced that allow for a minimal loss of parallelism during the computation of input scalar products. A ‘greedy’ optimization procedure applied to the neuron-processor assignment is shown to gain a further reduction of computation time getting close to the lower bound. Using these methods, a considerable speedup in comparison to sequential computation is achieved.
Similar content being viewed by others
References
V.Braitenberg, “Cortical architectonics: general and areal”, in M.A.B.Brazie and H.Petsche (eds) Architectonics of the Cerebral Cortex, pp. 443–465, Raven: New York, 1978.
V. Braitenberg and A. Schüz, “Cortex: Hohe Ordnung oder größtmögliches Durcheinander”, Spektr. d. Wiss., Mai 1989.
D.Hammerstrom and N.Nguyen, “A VLSI architecture for high-performance, low-cost, on-chip learning”, in Int. Joint Conf. on Neural Networks Vol. II, pp. 537–544, Elsevier: Amsterdam, 1990.
CNAPS/VME Board Reference Manual, Rel. 1.0, Adaptive Solutions, Inc., Beaverton, USA, 1993.
H.-U.Heiß, “Prozessorzuteilung in Parallelrechnern”, BI-Wissenschaft: Mannheim, Wien, Zürich, 1994.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Möller, R., Paschke, P. Simulation of cortex-like neural networks on a CNAPS SIMD neurocomputer. Neural Process Lett 4, 67–74 (1996). https://doi.org/10.1007/BF00420615
Issue Date:
DOI: https://doi.org/10.1007/BF00420615