5. Optimization for TCAD

This chapter deals with various optimization related tasks in the TCAD field. The optimization framework that is available in the TCAD environment of the Institute for Microelectronics is presented. Then, a local gradient based optimizer is compared to two global optimization strategies, namely genetic algorithms and simulated annealing. Finally, an approach that combines local and global strategies to achieve good performance and high accuracy at the same time is presented.

Subsections

• 5.1 Introduction
  • 5.1.1 Inverse Modeling
  • 5.1.2 Calibration of Models
  
  
• 5.2 Optimization Framework
  • Job Farming Facility
  
  
• 5.3 Gradient Based Optimization Methods
  • Levenberg-Marquardt Optimizer
  
  
• 5.4 Global Optimization Methods
  • 5.4.1 Genetic Optimization Algorithms
    • Data Representation
    • Genetic Reproduction
    • GENOPT
  • 5.4.2 Simulated Annealing Algorithms
    • SIMAN
  
  
• 5.5 Combination of Local and Global Methods
• 5.6 Examples
  • 5.6.1 Comparison of Local and Global Optimization Strategies
    • Local Optimizer
    • Genetic Optimizer
    • Simulated Annealing Optimizer
  • 5.6.2 Combination of Local and Global Optimization Strategies
    • 5.6.2.1 Optimizer Protocol
    • 5.6.2.2 Optimizer Interaction
    • 5.6.2.3 Results
  • 5.6.3 Calibration of a Model for Silicon Self-Interstitial Cluster Formation and
    Dissolution
    • 5.6.3.1 The Model for Cluster Formation and Dissolution
    • 5.6.3.2 Calibration Task

2003-03-27