Performance of a cascadable neural network VME-module with Intel 80170NX chips

C. S. Lindsey, Th Lindblad, J. R. Vollaro, G. Szèkely, J. Molnar

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

We describe an experimental research prototype VME-module with three commercially available Intel 80170NX analog neural network chips. Each chip can hold one layer with up to 64 neurons of a feedforward network and a single module can thus be configured for 1 to 3 layer networks. Multiple boards can be cascaded together to build broad structured networks. We discuss tests of a neural network application in high energy physics. Some supporting circuits are also discussed.

Original languageEnglish
Pages (from-to)466-471
Number of pages6
JournalNuclear Inst. and Methods in Physics Research, A
Volume351
Issue number2-3
DOIs
Publication statusPublished - Dec 1 1994

ASJC Scopus subject areas

  • Nuclear and High Energy Physics
  • Instrumentation

Fingerprint Dive into the research topics of 'Performance of a cascadable neural network VME-module with Intel 80170NX chips'. Together they form a unique fingerprint.

  • Cite this