A comparative study of complexity-based capacitance macro-models

We present a comparative study of several RT-level capacitance estimators that can be used in the more general framework of macro-modeling-based high-level power estimation. To enforce their independence from both structural and technological details, the various estimators we consider are based on different technology-independent quick-synthesis paradigms that do not require any interaction with the chosen technology library. They all rely on decomposition of the original description into abstract, lower-level primitives for which the characterization in terms of the target library can be performed once and for all.