Automation Of Thermal Model Extraction And Optimization

Abstract

Semiconductor device model optimization is highly beneficial; it improves the accuracy with which we can predict the performance of real devices and whole circuits before they are fabricated. This helps save time and money in being able to predict undesirable circuit responses. Although, it proves to be very useful, it can become very time consuming. The goal of this thesis is to use readily available tools to enhance and automate the somewhat complex process of optimization and streamline it so that results can be analyzed quicker with the goal of generating VBIC models that accurately match the behavior of the real device.