Of course, the (overemphasized) focus on methodology application
justifies the achievement of short-term results at the cost of
long-term properties. It is easily forgotten that many of
the difficulties TCAD is struggling with now were caused long ago when
early applications were designed as isolated, ``pioneer''
simulators
,
each with a different philosophy and often without the faintest
idea that they might someday be integrated into larger CAD
systems.
Looking at TCAD software, whether it is industrial or academic, it seems that everything is excused if just an urgent problem at hand can be solved. There are just very few, notable exceptions from the general unawareness of the software life cycle. A good characterization of this problematic situation has been given by Mar [14]
Because of the cyclical nature of technology development processes,
the time windows during which TCAD tools can impact a technology
design are relatively narrow. Technology development schedules are
driven by external market requirements and technology decisions will be
made with or without TCAD tools. Once specific technology decision
windows are passed the opportunity for applying TCAD to that phase of
the design are typically gone.
Application windows define what and when capabilities must be
delivered in order to be useful. Although this focus can cause
development to be focused on near term solutions, this has not
been the case at Intel where multiple generations of IC technologies
are developed concurrently.
To gain an understanding for potential TCAD problems, it is instrumental to recognize that Technology CAD is a continuous, dynamic process of creation and application of methodology rather than a static problem. This conception of TCAD is still not very common, which often leads to severe misunderstanding and entirely wrong expectations.