Algorithmic Optimization of Transistors Applied to Silicon LDMOS

We propose a pioneering approach that integrates optimization algorithms and technology computer-aided design to automatically optimize laterally-diffused metal-oxide-semiconductors (LDMOS) with field-oxide structure. define the ratio of square breakdown voltage divided by specific on-resistance as figure-of-merit (FOM) objective function our optimization. compare performance three different algorithms: Nelder-Mead, Powell, Bayesian Optimization. show how LDMOS evolves each reach their optimized straightforward Nelder-Mead leads local optimum when optimizing over six input parameters. find Optimization is most data-efficient method global structure in multi-domain space.