Title :
Considerations for neural network hardware implementations
Author :
Collins, Dean R. ; Penz, P. Andrew
Author_Institution :
Texas Instrum., Dallas, TX, USA
Abstract :
The requirements for neural network hardware possess unique fabrication difficulties for chip designers. Some of these difficulties involve the rapidly developing state of neural network theory. Another difficulty centers on the complexity of interconnections between neurons. Both semiconductor and optical implementation schemes are reviewed and examined with respect to practical implementation within the next several years
Keywords :
digital integrated circuits; integrated optoelectronics; linear integrated circuits; monolithic integrated circuits; neural nets; analogue devices; chip designers; digital devices; fabrication difficulties; neural network hardware implementations; neural network theory development; neuron interconnection complexity; optical implementation scheme review; practical implementation; semiconductor implementation scheme review; Backpropagation algorithms; Circuit testing; Computational modeling; Computer networks; Integrated circuit interconnections; Neural network hardware; Neural networks; Neurons; Power system interconnection; Random access memory;
Conference_Titel :
Circuits and Systems, 1989., IEEE International Symposium on
Conference_Location :
Portland, OR
DOI :
10.1109/ISCAS.1989.100480
