Stats

  • Locked Locked
  • Replies 2
  • Subscribers 90
  • Views 2763
  • 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

Generating min/max SDF files when in MMMC

archive

In previous projects I used multi-mode analysis
and CTE report mode (SOC52). I generate best and worst
case parasitc files for STA and also MIN and MAX
delay files (SDF). I used the following commands:

setAnalysisMode -setup
rcOut -worst -spf file_name_worst.spf
delayCal -sdf file_name_worst.sdf

setAnalysisMode -hold
rcOut -best -spf file_name_best.spf
delayCal -sdf file_name_best.sdf

But now I'm using SOC62 and multi-mode
multi-corner analysis and the above commands does
not work anymore. To generate parasitic files I'm
using now:
rcOut -rc_corner WORST -spf file_name_worst.spf
rcOut -rc_corner BEST -spf file_name_worst.spf

It seems be ok. Is this the right way to get
parasitic files in MMMC?

And how to get min/max delay files (sdf) in MMMC?
Usign the command delayCal, the resulting delay cells are
quite the same between MIN and MAX sdf files (less than 15%). I think there is no effect on
delay calculation when switching between hold and
setup analysisMode.
I tried use write_sdf but looking at the file
header, the process, voltage and temperature
values indicates that both sdf files are for
typical case (which is the default opCond).

Can anyone help me?

Regards,
Cristiano.


Originally posted in cdnusers.org by clsantos

  • Another interesting point:
    Using command "setDelayCalMode -feDc" the
    resulting delay cells are very different from
    before when I used signalStorm. Now there are
    a lot of difference between MIN and MAX delay
    cells in the generated sdf files.
    Why this difference? What is the recommended delay
    calculation mode?

    My felling is that when in signalStorm mode, delay
    calculation considers the opCond set by command
    setOpCond. As in my case I did not use this
    command, delay calculation are considering the
    default opCond (typical) and only switching
    between worst and best wire parasitic information
    in RC extracted database. What do you think about it?

    I attached two reports of the same path (setup),
    one using -signalStorm and other using -feDc.

    Regards,
    Cristiano.


    Originally posted in cdnusers.org by clsantos

  • In the 5.2 releases, SOCE was only  multi-mode, so the delay and extraction subsystems
    did not change.

    In 6.1 and 6.1, SOCE is also multi-corner.   In the releases, you should
    use write_sdf -view xyz  to dump the SDF file of interest.   Delays are
    subject to both the mode effect and the corner effect - so the control
    is at the view level.

    If the SDF header info is not correct - get a PCR filed.. this is mostly cosmetic.  If the values
    do not change that is a real  problem - assuming your delay corners are substantially
    different.



    Originally posted in cdnusers.org by ejm

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.