Originally, these tools were used as individual units to improve the equipment, methods and materials related to a particular process step. Today, the role of TCAD is much wider and includes the analysis and characterization of the devices themselves. For the active devices the semiconductor transport equations at certain operating conditions are solved, and for the passive/parasitic elements an appropriate extraction tool is used. By using simulation methods technology specialists can observe what is going on inside the devices (useful in the development phase) which is impossible with measurement techniques applied directly to real ones. Consequently, a reduction in the time and cost required to improve an existing technology or developing new ones is gained.
As technology becomes more complex, both in number of process steps and physics involved, the demands on the simulation capabilities are also increasing. Modern TCAD tools are put together in frameworks that simplify the integration of different applications by sharing common tasks, using uniform input/output mechanisms and a better management of the simulation data [11].
The reduction of minimum feature sizes makes the devices and structures more sensitive to process variation. Optimizations performed by hand or controlled by the user cannot achieve the optimum levels and therefore TCAD frameworks include the possibility of optimizing entire simulation flows. The goals can be defined by using Design of Experiments (DOE) and Response Surface Methodology (RSM) methods [12].
Another important topic in TCAD frameworks is easy model implementation. Frameworks must offer process and device engineers a simple model interface, as these need to update their models along with technology evolution. Complementary, methods for tool and model calibration should be provided, as well.