Challenges for predictive process simulation in sub 0.1 μm silicon devices

The promise of process simulation has always been that it could replace costly experiments in the development of new technology. The reality has been far less, primarily because of deficiencies in the physical models underlying such simulations. This situation is changing rapidly however, as better physically based models are developed. This paper describes recent progress in developing physically based process simulators like SUPREM. A number of examples will be shown which will illustrate the growing predictive power of such TCAD tools, even for silicon structures well beyond current manufacturing practice.