Stats

  • Locked Locked
  • Replies 7
  • Subscribers 125
  • Views 15006
  • 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

Running APS with VerilogA and bsource modules

fomin

In Virtuoso Accelerated Parallel Simulator User Gide I found the following statement:

When running APS, the VERILOGA and BSOURCE modules are forced to be compiled. Because of this, setting the value CDS_ADHLCMI_ENABLE NO has no effect. Also, the command line arguments -ahdlcom and -bsrccom don't produce any result.

Does that mean APS doesn't support VERILOGA and BSOURCE modules? Some foundries use BSOURCE for resistor models. (TSMC 0.18u, for example) 

 

  • No, it means that veriloga and bsource components are always compiled. Since the purpose of compilation is to make the simulation faster, it would be a bit odd to use APS (which is intended to be an optimised, fast, accurate simulator) and not take advantage of model compilation.

    I'm not sure how you could read this statement and come to the conclusion that veriloga and bsource are not supported!

    Regards,

    Andrew

  • Because I tried to run APS for two models and found out that for one model it gives up to 600% CPU usage, but for another no more than 100%. I checked model files and noticed that one of them that is running with 100% CPU usage used bsource for resistor models.

    After that I found this statement in User Guide and asked this question

  • Thanks for the clarification.

    I suggest you file a service request via http://support.cadence.com so we can investigate why the speed is low with that particular example.

    It sounds a bit strange to me...

    Regards,

    Andrew.

  • Everything is cleared now.

    I just didn't noticed the following message of the second task:

    "Multithreading is disabled due to the size of design being to small."

  • Even on smaller designs you can often get speed improvements with APS, even if it's just using a single core - so maybe the elapsed time is still shorter...

    Regards,

    Andrew.

  • But why not use different thread for different parameter set during parametric analysis. If I have small design but need to run long simulation time totally it will save much time.

  • There are already ways of doing that with Distributed Processing in ADE L (and ADE XL), and in fact those can work across machines so do not have to even be tied to the same machine. So it doesn't strictly need to be done in the simulator.

    That said, we are always looking at other ways of taking advantage of multi-threading, and this is one of a number of areas that might be considered for the future.

    Best 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.