Re: In models we trust

Passsive models should yield identical results in different simulators.
Only in active models (IBIS buffers) there may be differences due to interpolation/extrapolation schemes.

> From owner-ibis@server.vhdl.org Fri May 30 18:23 EDT 1997
> Received-Date: Fri, 30 May 1997 18:23:00 -0400
> From: apanella@molex.com
> X400-Received: by /PRMD=molex/ADMD=molexnet/C=us/;
> Relayed; 30 May 97 17:05 CDT
> X400-Received: by mta "molexmta" in /PRMD=molex/ADMD=molexnet/C=us/;
> Relayed; 30 May 97 17:13 CDT
> Priority: normal
> P1-Message-ID: us*molexnet*molex;0865030389/0000002181/1
> Original-Encoded-Information-Types: IA5-Text
> MIME-version: 1.0
> To: ibis-users@vhdl.org, owner-ibis@server.vhdl.org
> Subject: Re: In models we trust
> Content-Type: text
> Content-Length: 2529
> X-Lines: 72
>
> Hello all,
>
> Good comments Stephan!!!
>
> But it does bring up the question....
>
> Who verifies the performance of the simulation vendors products?
>
>
> My understanding (specifcally in regards to the potential connectors
> models), that it is possible for each simulation vendors product to
> handle passive models fdifferently.
>
> Therefore, a model I create for one IBIS model solver will look
> differently on othe IBIS solvers.
>
> I would not want to be required to make a different passive component
> (ie connector model) for each IBIS simulation vendor.
>
> Comments??
>
> Regards,
> _gus panella
> molex, incorporated
> apanella@molex.com
>
>
> ______________________________ Reply Separator _________________________________
> Subject: In models we trust
> Author: owner-ibis@server.vhdl.org at internet
> Date: 97/05/30 1:51 PM
>
>
> Hello Luke:
>
> Well, first off, I would never "blindly" trust ANY model -- spice,
> IBIS, or otherwise. Remember that all models are accurate only over
> a given environment/operating range, and a model may only be good
> for a specific kind of simulation. For example, a model created
> for signal integrety work is probably not approriate for doing EM
> simulations. Before using any model one must understand it's
> purpose and the conditions under which it gives accurate results.
>
> To answer your specific question -- the model vendor is responsible
> for verifing the data in, and results obtained from, an IBIS model. The IBIS
> forum does make sure that the model passes the golden parser before
> it's placed on our web site, but we don't have the capabilites to verify
> that the model gives "good results". If you have questions about a specific
> model you should contact the model supplier directly.
>
> Regards,
> Stephen Peters
> Intel Corp.
>
>
>
> Hi,
>
> When IC vendors create IBIS models for their components, how do they
> verify whether they are correct or not, compared to either the original
> SPICE models or to physical measurement data? Do they actually do this
> comparison? Is there a simulation methodology to do this comparison?
>
> I am just a user of IBIS models and will never create them. So I am
> mainly interested in knowing whether I can always (blindly) trust the
> models I get from the vendor Web sites or directly from the vendors.
>
> Thanks.
>
> Luke Chang
> Ascom Nexion, Inc.
> Acton, MA
>
 
Received on Fri May 30 15:36:21 1997