TCAD Augmented Machine Learning for Semiconductor Device Failure Troubleshooting and Reverse Engineering
Publication Date
9-1-2019
Document Type
Conference Proceeding
Publication Title
2019 International Conference on Simulation of Semiconductor Processes and Devices (SISPAD)
DOI
10.1109/SISPAD.2019.8870467
Abstract
In this paper, we show the possibility of using Technology Computer Aided Design (TCAD) to assist machine learning for semiconductor device failure trouble shooting and device reverse engineering. When TCAD simulation models and parameters are properly chosen and calibrated, large number of devices with random defects and structural characteristics can be generated and simulated. The results can then be used to train machine learning algorithms to predict the defect and structural characteristics of a device with given electrical characteristics (such as IV's and CV's). 1D PIN diode with various layer thicknesses and doping concentrations are used in this study. It is showed that with less than 2000 training samples, by using simple linear regression, one can achieve good prediction of layer thickness and doping of a given IV curve.
Keywords
Machine Learning, Reverse Engineering, Semiconductor Defects, TCAD
Department
Electrical Engineering
Recommended Citation
Y. S. Bankapalli and H. Y. Wong. "TCAD Augmented Machine Learning for Semiconductor Device Failure Troubleshooting and Reverse Engineering" 2019 International Conference on Simulation of Semiconductor Processes and Devices (SISPAD) (2019). https://doi.org/10.1109/SISPAD.2019.8870467
