scispace - formally typeset
Search or ask a question

Showing papers by "Timo Hämäläinen published in 1996"



Journal ArticleDOI
TL;DR: The tree shape parallel computer system, GAPA (Genetic Algorithm Parallel Accelerator) with special hardware for GA computation, is described in detail and mappings for centralized and distributed GA models are given and their performance has been measured for different population sizes.

9 citations


Book ChapterDOI
16 Jul 1996
TL;DR: The architecture of linearly expandable partial tree shape neurocomputer (PARNEU) is presented andappings of algorithms are presented for Hopfield and perceptron networks, Sparse Distributed Memory, and Self-Organizing Map.
Abstract: The architecture of linearly expandable partial tree shape neurocomputer (PARNEU) is presented. The system is designed for efficient, general-purpose artificial neural network computations utilizing parallel processing. Linear expandability is due to modular architecture, which combines bus, ring and tree topologies. Mappings of algorithms are presented for Hopfield and perceptron networks, Sparse Distributed Memory, and Self-Organizing Map. Performance is discussed with figures of computational complexity.

5 citations


Book ChapterDOI
16 Jul 1996
TL;DR: Mapping of multilayer perceptron networks are presented for a parallel neurocomputer system called TUTNC (Tampere University of Technology Neural Computer) and an example mapping is given for error backpropagation training utilizing combined neuron and weight parallel schemes.
Abstract: Mapping of multilayer perceptron networks are presented for a parallel neurocomputer system called TUTNC (Tampere University of Technology Neural Computer). The tree shape architecture of the system offers implementation possibilities at several levels of parallelism for both forward mode of operation and network training. An example mapping is given for error backpropagation training utilizing combined neuron and weight parallel schemes. The performance is measured for some example network sizes, and estimations are given for a larger system.

5 citations


Journal ArticleDOI
TL;DR: A parallel algorithm which trains artificial neural networks on the TUTNC (Tampere University of Technology NeuroComputer) parallel tree shape neurocomputer and it is possible to add more processing units to the system without any change in the number of communication transactions required.

3 citations


Proceedings ArticleDOI
04 Sep 1996
TL;DR: Two different realizations of a sparse distributed memory (SDM) model are presented and it is shown, that SDM can be well parallelized in TUTNC.
Abstract: This paper presents two different realizations of a sparse distributed memory (SDM) model. For parallelization purposes, addressing, storage and retrieval operations are explained in detail and some existing implementations in various computing platforms are considered before introducing the tree shaped parallel computer, TUTNC (Tampere University of Technology Neural Computer). The architecture and main features of TUTNC are presented in order to map SDM to the system in columnwise and rowwise manner. Mappings are compared in terms of measured execution time with different parameter sets. Speedup and performance estimations are also given for a larger system. The results show, that SDM can be well parallelized in TUTNC.

2 citations


Proceedings ArticleDOI
05 Aug 1996
TL;DR: A secure infrared remote control system that matches several data encoding methods and carrier modulation techniques and is suitable for electronic door locking purposes for domestic and automotive applications.
Abstract: A secure infrared remote control system is presented. Offering flexible programmability, the presented system matches several data encoding methods and carrier modulation techniques. Arbitrary long data codes can be transmitted within the capacity of the system memory. The low cost of the implementation is due to commercial components and a minimum number of circuits on the small sized printed circuit board. Powered by a battery energy source, the presented system is suitable for electronic door locking purposes for domestic and automotive applications. An example application for a car security access control is presented. The system has been tested under variable environment conditions and found to be very reliable and robust.

1 citations