A novel methodology for neural compact modeling based on knowledge transfer. (December 2022)

• Record Type: Journal Article
• Authors: Cha, Ye Sle; Park, Junghwan; Park, Chanwoo; Chong, Soogine; Kim, Chul-Heung; Lee, Chang-Sub; Jeong, Intae; Cho, Hyunbo
• Is Part Of: Solid-state electronics. Volume 198(2022)
• Publication Date: 2022-12
• Subjects: Artificial neural network -- Compact modeling -- Deep learning -- Knowledge transfer -- Meta learning -- MOSFET -- Statistical modeling -- Transfer learning; Semiconductors -- Periodicals; Semiconducteurs -- Périodiques; 621.38152
• DOI: 10.1016/j.sse.2022.108450
• Languages: English
• ISSNs: 0038-1101

Abstract

Abstract: This work presents a novel approach of using knowledge transfer to increase the accuracy of artificial neural network (ANN)-based device compact models, or neural compact models. This is useful when the amount of data available for training an ANN is limited. By utilizing relatively abundant data of a previous technology node, physical phenomena that are not evident in the limited data of the target technology node (e.g. gate-induced drain leakage) are accurately predicted. When meta learning algorithms are used, the accuracy of the model significantly increases, with relative linear error 10 times lower compared to the case when prior knowledge is not incorporated. The proposed methodology can be used to model future generation devices with limited data, utilizing data from well-characterized past technology node devices. Highlights: We introduce neural compact models using knowledge transfer methods. The proposed models aim to tackle the scarcity of data for a target device. They show excellent accuracy, particularly when meta learning algorithms are used. Physical phenomena such as GIDL are also accurately predicted.