[19a-Z32-1~10] 23.1 Joint Session N "Informatics"

△ [19a-Z32-8] Construction of characteristics prediction model for diamond semiconductor field effect transistors

〇Arisa Nishibu¹, Hachiya Ryouta², Fujii Mami¹, Kutsukake Kentaro^2,3, Ujihara Toru², Uraoka Yukiharu¹ (1.NAIST, 2.Nagoya Univ., 3.RIKEN AIP)

Keywords:Diamond, Semiconductor, Machine learning

Diamond semiconductors are said to be the ultimate semiconductor materials, and new devices are expected to be feasible, but current diamond semiconductors are expensive and take time to manufacture. In this study, a simulation was carried out considering the parameters representing the material properties of diamond. Furthermore, by learning using a neural network and optimizing the combination of parameter values that give arbitrary electrical characteristics, it was estimated in 1/500 of the time required by the conventional method.