Enabling process control though predictive design and virtual metrology for high product mix manufacturing

The semiconductor foundry industry faces challenges in managing diverse customer demands and complex manufacturing processes. Variations in the chemical vapor deposition process affect transistor parameters and yield. Siemens' Calibre (R) software with machine learning techniques create a virtual metrology model that outperforms traditional methods. An advanced process control system, incorporating design features and real-time data, improves process capability and reduces film thickness variations in high-mix product foundry fabs, as confirmed by control simulations.