Reliability Prediction for Automotive 5nm and 7nm Technology node by using Machine Learning based Solution

Automotive semiconductor products demand high reliability requirements. Thus, only electrical test after fab-out may not be sufficient for a reliability management. This paper proposes a machine learning (ML) based solution to predict reliability in real time during the middle of wafer processing. The solution includes fab-data augmentation (FDA) to estimate missing values by using partially available measurement data. This is followed by building a ML model to identify and rank the critical process steps that impact reliability. ML model is also used to predict reliability of wafers to screen low reliability wafers out in advance during the fab process by real time monitoring.