A Physics-Informed Automatic Neural Network Generation Framework for Emerging Device Modeling
Micromachines · 29.05.2023 · 14 (6) · 1150
DOI: 10.3390/mi14061150 ↗Kurz gefasst
Fachartikel in der Zeitschrift Micromachines, online erschienen am 29.05.2023. Verfasst von Guangxin Guo und 4 weiteren Autor:innen. Beteiligte Einrichtung: School of Microelectronics, Xidian University, Xi'an 710071, China. Die Arbeit wurde bisher 17-mal zitiert (Crossref) und nennt 28 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: China National Key R&D Program, Project of Science and Technology on 294 Reliability Physics and Application Technology of Electronic Component Laboratory, Xidian University.
Abstract (Originalsprache)
With the rapid development of semiconductor technology, traditional equation-based modeling faces challenges in accuracy and development time. To overcome these limitations, neural network (NN)-based modeling methods have been proposed. However, the NN-based compact model encounters two major issues. Firstly, it exhibits unphysical behaviors such as un-smoothness and non-monotonicity, which hinder its practical use. Secondly, finding an appropriate NN structure with high accuracy requires expertise and is time-consuming. In this paper, we propose an Automatic Physical-Informed Neural Network (AutoPINN) generation framework to solve these challenges. The framework consists of two parts: the Physics-Informed Neural Network (PINN) and the two-step Automatic Neural Network (AutoNN). The PINN is introduced to resolve unphysical issues by incorporating physical information. The AutoNN assists the PINN in automatically determining an optimal structure without human involvement. We evaluate the proposed AutoPINN framework on the gate-all-around transistor device. The results demonstrate that AutoPINN achieves an error of less than 0.05%. The generalization of our NN is promising, as validated by the test error and the loss landscape. The results demonstrate smoothness in high-order derivatives, and the monotonicity can be well-preserved. We believe that this work has the potential to accelerate the development and simulation process of emerging devices.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- School of Microelectronics, Xidian University, Xi'an 710071, China.
ORCID-Profile: Guangxin Guo, Hailong You, Cong Li, Zhengguang Tang, Ouwen Li
Förderung
- China National Key R&D Program
- Project of Science and Technology on 294 Reliability Physics and Application Technology of Electronic Component Laboratory
- Xidian University
- Beijing Microelectronics Technology Institute
- Project of Science and Technology on Reliability Physics and Application Technology of Electronic Component Laboratory
- China National Key R&D Program
- 111 Project of China
- National Natural Science Foundation of China
Zitieren
APA
Guo, G., You, H., Li, C., Tang, Z., & Li, O. (2023). A Physics-Informed Automatic Neural Network Generation Framework for Emerging Device Modeling. Micromachines, 14(6), 1150. https://doi.org/10.3390/mi14061150
Vancouver
Guo G, You H, Li C, Tang Z, Li O. A Physics-Informed Automatic Neural Network Generation Framework for Emerging Device Modeling. Micromachines. 2023;14(6):1150. doi:10.3390/mi14061150
BibTeX
@article{Guo2023,
author = {Guo, Guangxin and You, Hailong and Li, Cong and Tang, Zhengguang and Li, Ouwen},
title = {A Physics-Informed Automatic Neural Network Generation Framework for Emerging Device Modeling},
journal = {Micromachines},
year = {2023},
volume = {14},
number = {6},
pages = {1150},
doi = {10.3390/mi14061150}
}
EUVIMED