Semiconductor fabs have historically used inspection and metrology equipment for yield learning – to monitor individual process tools for excursions and to assure that their manufacturing processes stay in control to maximize productivity. The increasing semiconductor content in automobiles, driven by growth in ADAS, electrification and autonomy, has put a growing focus on the quality and reliability of these devices and their implications for consumer safety and satisfaction.
This push for better reliability is embodied in the industry’s Zero Defect initiative, as car manufacturers drive semiconductor DPPM requirements for their chip suppliers below 1PPM. More than half of the semiconductor failures that occur on the automotive assembly line today (so-called 0km failures) have their origin in semiconductor fab defectivity. To successfully meet their customers’ reliability goals, fabs across the full range of design rules and device types are looking at their inspection and metrology data in a fresh light. Not only are they using their process control tools to do a better job of reducing overall sources of defectivity, they are also looking to the data to help disposition individual die, removing high-risk defective die from the supply chain.
In this keynote, KLA will update some of these industry trends and some of the exciting steps forward in quality and reliability made possible by these novel methods.