I would ask you which is more appreciated in IC design, the Montecarlo or Worst case analyses ??
for me I see that worst case analyses is more reliable and consuming less simulation time, any way I would like to share your opinions
With MonteCarlo analysis you can find runaway cases which you wouldn't find with w.c. analyses, see e.g. the following cross (or quiescent) output current curves over the full input voltage range of a class AB buffer opAmp with incorrectly biased AB output stage:
The PVT w.c. corner analysis had produced a curve like that in the middle (about 25µA).
We also found an oscillator-not-starting case by MC analysis for a parameter combination far from the corners - later confirmed by silicon.
Last not least you can get a yield estimation, if you combine local & global variation distributions, see this tutorial:
Thank you erikl for your reply
I agree with you but I think if the system run after the WC analyses then it mean it will run by any circumstances, and this what we want from the circuit