Deep Sub-micron Unit Process Parameter Extraction And Modeling Techniques

Cell characterization and modeling are very complex process in Very Deep Submicron(VDSM) technologies. It exists great science significance and very high practical value to study cell characterization and modeling. The dissertation presents an automatic flow to characterize and model cells accurately and quickly by the stimulus reduction algorithm, the input signal driving algorithm, the characterization point dynamic insertion algorithm, the result table optimization algorithm and input signal integrity analysis algorithm, which reaches the advanced level in this field.The dissertation proposes the stimulus generation algorithm for cell characterization. Based on the function analysis of cells and referred to multi-path sensitization algorithm, the algorithm provides logic stimuli for cell characterization. For the case of many pins, however, the algorithm reduces the stimuli by analyzing the charging/discharging status of the circuit internal nodes.The input signal-driving algorithm for VDSM is given in the dissertation. The algorithm can quickly and accurately acquire the input signal, which is consistent with the real signal.The dissertation explains the algorithm to ascertain the cell working range and to choose characterization points. For the cell working range ascertainment, the algorithm determines the maximum output load, the minimum output load, the maximum input signal slope and the minimum input signal slope; For the cell working surface representation, the algorithm inserts the characterization points dynamically.The table optimization algorithm for redundantly characterized points is proposed. Its result table can represent the whole parameter surface accurately and concisely.Finally, the timing effect of the signal integrity problem is discussed. The dissertation analyses the timing effect of IR drop, and gives the expression for the timing effect; and the dissertation analyses the crosstalk effect of cell internal routing, presents a cell internal routing algorithm for crosstalk minimization and a crosstalk model for timing analysis.