EsteMate: a tool for automated power and area estimation in analog top-down design and synthesis

Abstract

A novel methodology to derive power and area is presented. The method consists of the generation of a training set and a subsequent training of an Artificial Neural Network. The resulting estimators do not only predict power and area accurately and efficiently, they also reflect the complex interaction between different specifications. The method has been implemented in a tool, EsteMate, and is illustrated with practical examples.