Re: [SI-LIST] : IBIS Model Verification (correction)

Chris Rokusek (crokusek@viewlogic.com)
Thu, 10 Sep 1998 14:37:16 -0700

Ray,

I really like your comment "...the model developer should have ensured
model compliance prior to releasing to the user community..."

Unfortuntely, it seems that many IBIS models end up out there with
severe problems. What really complicates the matter is when an IC
vendor is unwilling to release the model (EVEN UNDER NDA) to EDA
vendor(s).

This practice creates much more work for all parties involved--
customers, IC vendors, EDA vendors. It makes IC vendors and EDA vendors
look bad to the customer as well as create doubt in the IBIS standard
itself. It also forces the EDA vendor to diagnose the problems through
the eyes of the customer which is painful and error prone and often the
"right" correction is ambiguous (Try and fix VI curves w/o looking at
them individually as well as the appropriate sums).

Of course we try to automate the severe fixes within the software at
run-time but we end up trading robustness for accuracy for no good
reason.

If IC vendors begin to use the IBIS model review committee (or at least
under NDA) we'll all save some time.

Chris Rokusek
Viewlogic Systems

Ray Anderson wrote:
>
> So what if a user receives a commercial model that he/she believes is
> defective in some way and notifies the model's developer of the fact?
> Then the developer submits it for verification. Will this overload the
> system? This would provide a single point of contact between the
> developer and the verifier.
>
> Granted, the model developer should have ensured model compliance
> prior to releasing it to the user community, but at least this would/could
> provide verification after the fact in problematic cases.
>
>
> Ray Anderson
> Sun Microsystems Inc.
> >
> >
> > Hi Weston, Stephen,
> >
> > I need to modify a statement made by Stephen.
> > The "Free" IBIS model review is NOT for CAE users. It was
> > setup to assist model developers at the IC house with their
> > models. We DO NOT have the bandwidth to check models submitted
> > by users of all the CAE tools in the world.
> > We DO have the bandwidth to check models submitted by the IC houses.
> >
> > If we had more volunteers we could expand this program into users.
> > I suggest a user network be created powered by other users that
> > volunteer. The "IBIS open user forum"
> >
> >
>
> **** To unsubscribe from si-list: send e-mail to majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at

**** To unsubscribe from si-list: send e-mail to majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****