Silicon compilation of neural networks

Author

Nigri, M. ; Treleaven, P. ; Vellasco, M.

Author_Institution

Dept. of Comput. Sci., Univ. Coll. London, UK

fYear

1991

fDate

13-16 May 1991

Firstpage

541

Lastpage

546

Abstract

The aim of PYGMALION, a neural network programming environment, is to fully integrate the silicon compiler with the programming environment. Neural networks written in PYGMALION´s nC network specification language translate either to binary (for simulation) or silicon (for hardware). The silicon compilation project is described. It comprises three major parts: the PYGMALION programming environment, including nC; the so-called generic-neuron VLSI architecture at the heart of the silicon compiler; and the strategy for compiling from nC to silicon, focusing on the high-level synthesis

Keywords

VLSI; circuit layout CAD; neural nets; programming environments; PYGMALION; generic-neuron VLSI architecture; high-level synthesis; nC network specification language; neural network programming environment; programming environment; silicon compiler; Circuit simulation; Computational modeling; Computer networks; Computer simulation; Neural network hardware; Neural networks; Programming environments; Silicon compiler; Specification languages; Very large scale integration;

fLanguage

English

Publisher

ieee

Conference_Titel

CompEuro '91. Advanced Computer Technology, Reliable Systems and Applications. 5th Annual European Computer Conference. Proceedings.

Conference_Location

Bologna

Print_ISBN

0-8186-2141-9

Type

conf

DOI

10.1109/CMPEUR.1991.257444

Filename

257444