Stats

  • Locked Locked
  • Replies 9
  • Subscribers 63
  • Views 4107
  • 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

PSS & PAC simulation running very slowly when running through OCEAN script in IC616

RFStuff

Dear All,

Pss & PAC simulation running very slowly when running through OCEAN script. But if I run in ADE it runs much faster.

I have filed a service request since almost 2 month.

But they say that the research team is looking into this.

Have anybody faced this issue. ? Is there any work-around for this ?

Kind Regards,

  • Please provide the case number so I can take a look. This should not be the case - and it is surprising. The only simple thing that can sometimes make a difference is that ocean echoes all the log file info to the output - and this can be a bit slow - but not usually enough to slow things down significantly.

    Other than that, there presumably is a difference in the netlist in both cases. If it was me looking into this, I would do a diff between the input.scs I get with ADE and that which I get from OCEAN.

    Regards,

    Andrew

  •  Dear Andrew,

    Thanks for your reply.

    The case number is 45785605.

    Ocean log shows the same time as ADE. But actually,  it takes more time. This was verified by the person handling the SR.

    Kind Regards,

  • This is indeed due to how OCEAN displays the simulator output into the CIW which is very slow - ADE does not do that (you get a fast log file window). I ran your testcase, and I see that the PAC in particularly is producing a lot of output in the CIW.

    If I add:  ?annotate "title" to both the analysis('pss ...) and analysis('pac ...) statements in the OCEAN script, this signficantly reduces the output and that speeds up the run to be very close to ADE's elapsed time.

    Regards,

    Andrew.

  • Hi Andrew,

    Thanks a lot for giving your valuable time to the case.

    Could you please confirm:-

    I should add following lines in my OCEAN scripts

    analysis('pac ?outType "Voltage"  ?sweeptype "absolute"  ?start "4G-100M"  
            ?stop "4G-0.1e6"  ?step "0.1M"  ?maxsideband "20"   ?annotate "title"  )


    analysis('pss ?fund "4G"  ?harms "4"  ?errpreset "moderate"  
            ?tstab "1u"  ?saveinit "yes"  ?method "gear2only"  ?tstabmethod "gear2only"  
            ?maxacfreq "100G"  ?annotate "title" )

    Kind Regards,

  • Hi Andrew,

    I did with ?annotate "title" option.

    The OCEAN scripts takes : 22.19 seconds

    The ADE simulation takes:- 19.22 seconds ( almost close)

    I think the above time difference is a FIXED one and should NOT increase with the long and complicated simulation .

    You may comment on the above.

    Without additions of ?annotate "title" option,

    The OCEAN scripts takes : 82.05 seconds.

    Though in IC5141 this problem was NOT there.

    Kind Regards,

  • The remaining time difference is due to the output that is still produced - there are ways to reduce that further but the benefit versus the effort is much lower - most of the output is coming from the pss and pac analyses. I agree, this needs to be fixed properly - my suggestions were merely a workaround for now. The remaining output is likely to be more static in size and so it's going to be mostly a fixed offset (although there are still some analysis-specific bits of info in the output). Andrew

  • If using IC616 ISR6 or later, there's an even better solution available. Put this at the beginning of your OCEAN script:

    envSetVal( "asimenv.misc" "includeSimLogInOCEAN" 'boolean nil)

    You then don't need to add the ?annotate "title" to each analysis statement. What this will do is suppress the echoing of the simulator output log in the CIW (or in the "ocean" output) whilst still writing everything to the spectre.out log file (just as ADE does) - and the elapsed time should be just as it is in ADE because you're not paying any penalty for printing in the CIW.

    I've written this up as a solution.

    Regards,

    Andrew.

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.