The accuracy problem for high-stack cells, as reported in[10], results from the fact that these cell-based models donot capture the staggered change in the states of stackedtransistors. For example, consider a 3-input Nand gate asshown in Figure 1, with a rising transition at pin A3. Initially, there is no current flow in the N-stack except for asmall leakage current. When A3 rises, transistor M3 turnson and starts to discharge node N2. However, due to Millercapacitance between A3 and N2, transistor M2 is turned offand does not conduct current until node N2 is dischargedbelow Vdd − Vt. Similarly, transistor M1 will start discharging node X when N1 is discharged enough to turn on M1.