Difference between Corners and Monte Carlo Analysis

Status
Not open for further replies.
Corner analysis uses special combination sets of conditions for PVT, e.g.
process=typ, V=nom and T=25°C for typ
process=worst, V=nom-10%, T=85°C for ind. w.c.
process=best, V=nom+10%, T=-40°C for ind. b.c.

Monte Carlo Analysis uses randomly selected condition sets in between, often selectable with limited conditions, e.g. with or without corner process conditions ("mismatch only"), or only process variations ("process only").
 

But in the corner analysis, what is done is a variation for example in the transistor dimensions? (let's suppose that we keep the temperature and voltage constant).

I've read this: "Then the corners cbest and cworst were created to model the smallest and largest cross sections that are in the allowed process variation." What is meant by allowed process variation?
 

AMSA84 said:
... in the corner analysis, what is done is a variation for example in the transistor dimensions? (let's suppose that we keep the temperature and voltage constant).
Click to expand...
Yes, but not only transistor dimensions.

AMSA84 said:
What is meant by allowed process variation?
Click to expand...
I think a better expression would be: process variation to be considered. You - as a customer - must allow for ;-)

Process variation includes variations of

- lateral dimensions (lithography, under-diffusion)
- implant dose
- junction depth
- gate oxide thickness and interface quality

and depends on the quality of process control.

All these variations affect the device parameters. The process corner limits should include the impact of process variation on the device parameters for a certain yield (e.g. 1σ or 3σ of total devices processed - to be accounted for).
 

Status
Not open for further replies.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…