Re: gEDA-user: gnucap: Multi-disciplinary / mixed language simulation

[Peter Clifton <pcjc2@xxxxxxxxx>]

On Tue, 2007-01-30 at 23:47 -0500, al davis wrote:
> Probably the best way is to write the model in Verilog-AMS.  
> Work is in progress to add that, and convert to Verilog-AMS as 
> the native language.

Are there any free tools to do Verilog-AMS currently? Are Verilog-A and
Verilog-AMS significantly different?

> Get the development snapshot of gnucap.  It supports plug-ins.  
> I do not recommend doing any models as "built-in" to gnucap.  
> Instead, do them as plugins.  Before the next stable release, 
> all existing models will move to plugins, so the core will have 
> no models.

Thats the code I've downloaded. Certainly a plugin based system is a
good idea for this type of work.

> If you want to write a gnucap model, use the model compiler.  It 
> takes care of the internals, so you can do it as a mix of 
> circuit level and mathematical level.  You can make a ".model" 
> file.  The model compiler generates C++, compile that into a 
> shared-object file (.so).  Then you can attach it at run time 
> using the "attach" command, or the "-a" command line option.  
> You get all of the benefits of a built-in model, with the 
> advantage that you can maintain it separately, and keep 
> complete control.

I've had a look at the .model files (Actually, I read your technical
documentation first... the linear algebra is a little over my head, but
I basically "get" how it works.)

> Since you mentioned octave ...  You could make a plug-in that 
> makes an interface to octave.

Which way around?

An octave plugin to:

a) Control gnucap
   - batch based - like the Ruby plugin I saw?
   or
   - integrated, so octave can be used as part of the simulation steps?

b) Import data from gnucap into an octave data-structure

c) Write gnucap control files from octave


OR:

A plugin to gnucap which:

a) Can write octave's data file format(s)


I'm actually avoiding octave as much as I can.. its maths libaries are
nice, but as an interpreted language, doing real tasks in it (if they
have loops) is really slow.

Since I inherited some Matlab code, I am using octave - it makes great
glue for trying things out, but I resort to C/C++ when I hit
bottlenecks.

Numpy / Scipy have been suggested as an alternative. (For the glue).

Thanks,

Peter C



_______________________________________________
geda-user mailing list
geda-user@xxxxxxxxxxxxxx
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user