Statistical process simulation for CAD/CAM

The authors formulate necessary conditions for predictive process simulation: physically-based models and tuning to manufacturing data, which are essential for CAM (computer-aided manufacturing) applications. They illustrate these predictive capabilities by applying the statistically based IC fabrication simulator FABRICS to the outcome of an industrial, scaled-down CMOS fabrication process. From the accuracy of the results, it is evident that the process simulation system based on physical models can be tuned to a desired accuracy.<>