TUTNC: a general purpose parallel computer for neural network computations

Abstract This paper presents the architecture and realization of the TUTNC (Tampere University of Technology Neural Computer) system. TUTNC is designed for parallel computation and it is suitable for several artificial neural network (ANN) algorithms. For parallel processing a MIMD multiprocessor architecture is chosen. Processors are connected via a tree shaped communication network, which can be programmed to perform efficient global operations. The architecture is modular and expandable. The cost of the system is kept low by using commercial DSPs (digital signal processors) and FPGA (field programmable gate array) chips. A small-scale prototype of four processors has been built. Performance is measured using some neural network mappings and estimations are also given for a larger system. Results show that a good cost/performance ratio can be achieved with this architecture.

[1]  Dan Hammerstrom.  A Highly Parallel Digital Architecture for Neural Network Emulation , 1991 .

[2]  John R. Nickolls,et al.  The design of the MasPar MP-1: a cost effective massively parallel computer , 1990, Digest of Papers Compcon Spring '90. Thirty-Fifth IEEE Computer Society International Conference on Intellectual Leverage.

[3]  Reinhard Männer,et al.  Multiprocessor And Memory Architecture Of The Neurocomputer Synapse-1 , 1993, Int. J. Neural Syst..

[4]  Kimmo Kaski,et al.  Some IC Implementations of Artificial Neural Networks Using Synchronous Pulse-Density Modulation Technique , 1991, Int. J. Neural Syst..

[5]  José G. Delgado-Frias,et al.  VLSI for Artificial Intelligence and Neural Networks , 1991, Springer US.

[6]  J.J.M. van Teeffelen.  Interfacing neural network chips with a personal computer , 1993 .

[7]  R. Hecht-Nielsen,et al.  Neurocomputing: picking the human brain , 1988, IEEE Spectrum.

[8]  Jukka Saarinen,et al.  Mapping of Some Neural Network Algorithms to a General Purpose Parallel Neurocomputer , 1993 .

[9]  H. T. Kung,et al.  The Warp Computer: Architecture, Implementation, and Performance , 1987, IEEE Transactions on Computers.

[10]  Jeff A. Bilmes,et al.  The Ring Array Processor: A Multiprocessing Peripheral for Connection Applications , 1992, J. Parallel Distributed Comput..

[11]  Colin Whitby-Strevens.  Transputers-past, present and future , 1990, IEEE Micro.

[12]  David M. Skapura,et al.  Neural networks - algorithms, applications, and programming techniques , 1991, Computation and neural systems series.

[13]  Tomas Nordström,et al.  Designing and using massively parallel computers for artificial neural networks , 1991 .

[14]  Angel L. DeCegama,et al.  The technology of parallel processing: parallel processing architectures and VLSI hardware (vol. 1) , 1989 .

[15]  Jukka Saarinen,et al.  Neural network computation in a parallel multiprocessor architecture , 1993, Proceedings of 1993 International Conference on Neural Networks (IJCNN-93-Nagoya, Japan).