Variability at the 32nm nodes in subthreshold regime

Variability at 32nm node in subthreshold regime

Greetings,

Can threshold voltage and gate length be varied concurrently ?Is it possible to consider simultaneous 3 sigma variations for both these parameters during modeling? Or these two parameters correlated (correlation coefficient must be considered)?

Re: Variability at 32nm node in subthreshold regime

(Which simulator are you using?)

Are you seeking to develop your own model, or do you want to run a monte carlo analysis on a given model and want to keep track of thos results where the threshold voltages are dependent on the gate length?

* If you have access to the models, you can do almost anything when defining what randomized variables should control which parameters. In spectre (cadence) that is pretty straight-forward. Go to the model card and look for e.g. "mc" or "mis*stat*blahblah" and you will see how those parameters are defined and you would be able to alter them accordingly. Probably, in this case, you want to set the correlation to 1.

* If you do not have access to the models, then I think it sounds more like a measurement task rather than a task to generate new random values. May I ask what you want to investigate?