Lu, AlbertMarshall, JordanWang, YifanXiao, MingZhang, YuhaoWong, Hiu Yung2023-05-022023-05-022022-12108468http://hdl.handle.net/10919/114896In this paper, two methodologies are used to speed up the maximization of the breakdown voltage (BV) of a vertical GaN diode that has a theoretical maximum BV of -2100 V. Firstly, we demonstrated a 5X faster accurate simulation method in Technology Computer-Aided-Design (TCAD). This allows us to find 50 % more numbers of high BV (>1400 V) designs at a given simulation time. Secondly, a machine learning (ML) model is developed using TCAD-generated data and used as a surrogate model for differential evolution optimization. It can inversely design an out-of-the-training-range structure with BV as high as 1887 V (89 % of the ideal case) compared to -1100 V designed with human domain expertise.application/pdfenCreative Commons Attribution 4.0 InternationalPower electronicsPower deviceBreakdown voltageDifferential evolutionGallium nitride (GaN)Machine learningTechnology Computer-Aided Design (TCAD)DiodeArticle - Refereedhttps://doi.org/10.1016/j.sse.2022.1084681981879-2405