Stats

  • Locked Locked
  • Replies 2
  • Subscribers 125
  • Views 14239
  • Members are here 0
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

monte carlo simulation

Nathan Chen

Hi, 

For monte carlo simulation, all devices are treated as uncorrelated.

there is constraint manager in IC61x to set constraint coefficient between devices, for example OP input stages.

is there simulation flow to import constraint coefficient into monte carlo simulation? 

thanks.

BR,

-nathan

Parents

  • Hi Nathan,

    If you use ADE XL/Assembler it will automatically take the correlation constraints and use them in Monte Carlo.

    However, I rather doubt this is useful. As I've said in a couple of threads recently (most recently Does MC mismatch simulation reflect best or arbitrary layout ?) foundry Monte Carlo mismatch models are modelling the remaining local random variation and not systematic mismatch caused by poor alignment. So I'd question the value in applying a correlation coefficient. Even if you wanted to, often the random parameters are quite abstract without obvious physical meaning and they are then combined with various equations into the device models, so it's hard to know which parameters you should correlate or what the correlation coefficients should be.

    Andrew

Reply

  • Hi Nathan,

    If you use ADE XL/Assembler it will automatically take the correlation constraints and use them in Monte Carlo.

    However, I rather doubt this is useful. As I've said in a couple of threads recently (most recently Does MC mismatch simulation reflect best or arbitrary layout ?) foundry Monte Carlo mismatch models are modelling the remaining local random variation and not systematic mismatch caused by poor alignment. So I'd question the value in applying a correlation coefficient. Even if you wanted to, often the random parameters are quite abstract without obvious physical meaning and they are then combined with various equations into the device models, so it's hard to know which parameters you should correlate or what the correlation coefficients should be.

    Andrew

Children
  • in reply to Andrew Beckett

    To my understanding, the foundries generally try to use uncorrelated statistical variables for their models. Correlations between model parameters are then realized by using the same statistical variable in the equations for several model parameters.

    Regarding correlations between devices, as Andrew already said, the mismatch models are already for a good layout (devices placed next to each other, with dummies, etc.), because the mismatch parameters are normally extracted from a layout that uses best practices to minimize mismatch.

Community Guidelines

The Cadence Design Communities support Cadence users and technologists interacting to exchange ideas, news, technical information, and best practices to solve problems and get the most from Cadence technology. The community is open to everyone, and to provide the most value, we require participants to follow our Community Guidelines that facilitate a quality exchange of ideas and information. By accessing, contributing, using or downloading any materials from the site, you agree to be bound by the full Community Guidelines.