With the continuing scaling of the feature size of integrated circuits, the relative variations of process parameters increase. This makes the traditional static timing analysis too pessimistic, because only the extreme values of parameters are considered. In statistical timing analysis, these parameters are directly modeled as random variables, with the correlation between them correctly handled. In this thesis the application of statistical timing analysis in the framework of hierarchical design of digital circuits is investigated. A method to extract timing models for combinational and sequential circuits is proposed. Additionally, this method can handle the correlation between modules in the hierarchical design correctly.
«
With the continuing scaling of the feature size of integrated circuits, the relative variations of process parameters increase. This makes the traditional static timing analysis too pessimistic, because only the extreme values of parameters are considered. In statistical timing analysis, these parameters are directly modeled as random variables, with the correlation between them correctly handled. In this thesis the application of statistical timing analysis in the framework of hierarchical desi...
»