Authors: J. Watts, C. Bittner, D. Heaberlin and J. Hoffman
Affilation: IBM Microelectronics Division, United States
Pages: 176 - 179
Keywords: genetic algorithms, compact model parameter extraction, BSIM3, PGAPack
Extracting an optimal set of parameter values for an FET device model is a complex problem. The final model must not only describe the performance of a set of hardware to an acceptable level of accuracy, but must satisfy criteria outside the deviceís allowed operating regime to ensure robust convergence properties during simulation. Traditional methods of parameter extraction which rely on gradient techniques can produce far-from-optimal solutions because of the presence of local optima in the solution space. As a result, parameter extraction has traditionally been more art than science, requiring several iterations by an experienced engineer. Genetic algorithms are well-suited for finding near-optimal solutions in irregular parameter spaces. We have applied a genetic algorithm to the problem of device model parameter extraction and are able to produce models of superior accuracy in much less time and with less reliance on human expertise.
Nanotech Conference Proceedings are now published in the TechConnect Briefs