Multiprocessor Realization of Neural Networks

Abstract

This research provides a foundation for implementing neural networks on multiprocessor systems in order to increase execution speeds and to accommodate more complex neural networks. The emphasis is on the use of affordable coarse grain multiprocessors to implement commercially available neural network simulators currently being run on single processor systems. A conceptual framework is presented based on the concepts of program decomposition, load balancing, communication overhead, and process synchronization. Four methodologies are then presented for optimizing execution times. A set of metrics is also introduced which make it possible to measure the performance enhancements over single processor systems, and analyze the effects of communications overhead, load balancing, and synchronization for various network decompositions. The application of these four methodologies to two neural network simulators on a multiprocessor computer system is discussed in detail. They are illustrated with practical implementations of networks ranging in size from six to twenty thousand connections. Two of the methodologies, the Pipeline and Hybrid approaches, exhibit speedups approaching the possible upper limits. The theoretical significant of this dissertation research is that is provides a basis for achieving efficient multiprocessor implementation of highly and massive neural networks. Traditionally, neural network research and development requires a considerable amount of time be spent in repeatedly evaluating and modifying network architecture and algorithms. As such, the engineering value of dissertation is that the time required to repeatedly execute networks in research and development can be significantly reduced.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Apr 01, 1990
Accession Number
ADA222146

Entities

People

  • Robert W. Bennington

Organizations

  • Air Force Institute of Technology

Tags

DTIC Thesaurus Topics

  • Artificial Intelligence
  • Artificial Intelligence Software
  • Brain
  • C Programming Language
  • Cognitive Science
  • Computer Languages
  • Computer Programming
  • Computer Programs
  • Computer Science
  • Computers
  • Content Addressable Memory
  • Information Systems
  • Network Architecture
  • Network Topology
  • Neural Networks
  • Operating Systems
  • Self Organizing Systems

Fields of Study

  • Computer science

Readers

  • Neural Network Machine Learning.
  • Parallel and Distributed Computing.
  • Software Engineering.

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks