An Analog Neural Computer with Modular Architecture for Real-Time Dynamic Computations
Penn collection
Degree type
Discipline
Subject
neuron
synapse
synaptic time constants
acoustical pattern analysis
Funder
Grant number
License
Copyright date
Distributor
Related resources
Author
Contributor
Abstract
The paper describes a multichip analog parallel neural network whose architecture, neuron characteristics, synaptic connections, and time constants are modifiable. The system has several important features, such as time constants for time-domain computations, interchangeable chips allowing a modifiable gross architecture, and expandability to any arbitrary size. Such an approach allows the exploration of different network architectures for a wide range of applications, in particular dynamic real-world computations. Four different modules (neuron, synapse, time constant, and switch units) have been designed and fabricated in a 2µm CMOS technology. About 100 of these modules have been assembled in a fully functional prototype neural computer. An integrated software package for setting the network configuration and characteristics, and monitoring the neuron outputs has been developed as well. The performance of the individual modules as well as the overall system response for several applications have been tested successfully. Results of a network for real-time decomposition of acoustical patterns will be discussed.