Dr. Dominik Thron
"Use of Big Data for Advanced Equipment Control in semiconductor manufacturing - Opportunities and Challenges"
Statistical Process Control Methods are well established at Semiconductor FABs. While the focus has been for years on results from Inline Metrology Tools - like Film Thickness, Critical Dimensions and Defectivity - it has now shifted towards Equipment Sensor Monitoring in recent times. That allows not only a faster detection of equipment failures, but also allows better process control with sensor data being available for each and every wafer. While this is an advantage it also increases the amount of data generated in Semiconductor FABs by orders of magnitude. This talk will discuss how GLOBALFOUNDRIES FAB1 is using the available data to extract the relevant signals today, what novel methods are be deployed to further increase the signal output and relevance, and what challenges we see by doing so.