I am using ADE-XL in IC6.1.5. The model file from the foundary defines two options for Mote-Carlo simulations. 1) lcoal mismatch only 2) global monte-carlo with local mismatch. I am using the second option to run Monte-Carlo where I want only to have process variations considered but not the local mismatch.
I setup Monte-Carlo for process variations only but it seems to me that mismatch is still considered. My simulation time went up by 10x as well as the memory usage (compared to Corners simulation time). The log file indicates that there are 33,323 mismatch instances though I only specifed process variation. I tried to go around the problem by specifying the main block to be excluded for mismatch but I got no improvment in simulation time nor in memory usage.
Could it be that the model file from the foundary which has global (process) and local(mismatch) overides the settings in ADE-XL and always run process with mismatch variations? Any idea how to enfore process only variations?
You probably should contact the PDK provider to find out how this is supposed to be used, or contact customer support so that we can look at the specific setup with you (it's hard to debug from a description without seeing the data).