Optimisation of Temperature Fields of Microsystems with Self-Organising Neural Nets

Thermal modelling and optimisation of parameter distributed systems is a rather time-consuming process. In this paper the problem of optimisation of temperature fields of VLSI circuits and systems is attacked by a selforganising neural net. The net directly solves the task generated by a heuristic a...

Full description

Saved in:
Bibliographic Details
Main Authors: P. Bratek, A. Kos
Format: Article
Language:English
Published: Wiley 2003-01-01
Series:Active and Passive Electronic Components
Online Access:http://dx.doi.org/10.1080/1042015031000073850
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Thermal modelling and optimisation of parameter distributed systems is a rather time-consuming process. In this paper the problem of optimisation of temperature fields of VLSI circuits and systems is attacked by a selforganising neural net. The net directly solves the task generated by a heuristic algorithm. No physical model of thermal phenomena is used. The proposed method is simple. Some examples and statistical results are presented. The proposed method is addressed mostly to large, high-speed system designs.
ISSN:0882-7516
1563-5031