Re: [SI-LIST] : HSPICE and XTK

About this list Date view Thread view Subject view Author view

From: Ray Anderson (raymonda@ha1mpk-mail.eng.sun.com)
Date: Fri Jul 28 2000 - 09:48:57 PDT


>
> Hello,
>
> I am currently looking to get myself some signal integrity software and I
> keep hearing wonderful things about Innoveda's XTK and bad things about
> Avanti! Hspice (see below). Is it really true? Is XTK really that wonderful?
> I will be simulating traces on boards, will XTK be able to successfully run
> a multi-board simulation? Will I have problems converting vendor's IBIS
> models into XTK format? What about translating board layout into lossy line
> models inside XTK? What about converting SPICE connector models in XTK
> format? Are there any caveats to using XTK?
>
> Thanks in advance to everyone.
>
> Mike Khusid
> Sitara Networks
>
>

Haven't used XTK so can't say how wonderful or not it is, however
your questioned is phrased to imply that the two tools (XTK and Hspice)
perform more or less equivalent functions.

Yes, they both can perform time domain analysis, but XTK is a special
purpose SI tool that can deal with layout databases, extract geometries
from them and perform many types of analysis'. It contains a pretty
nice field solver as I understand it.

Hspice, on the other hand, is a more general purpose tool. Basically spice is
spice. It is a simulator engine that can be used for many things, SI
simulation amongst them. It contains, in my opinion a rudimentary field
solver, that can be used to derive t-line parameters. Can't vouch for the
accuracy or lack of it as I usually use another field solver and then use
those results as input to some of the t-line models. Hspice can solve a whole
range of problems that XTK could never touch as well as ones that XTK excels
at. The thing is, it is up to the user to set up the simulation deck by
whatever means are available to feed to Hspice as input. With XTK, a lot of
the details of creating a simulation deck for analysis are handled by the
tool. Hspice's W element has some useful characteristics that make it most
appropriate for many SI simulations (the ability to simulate frequency
dependent loss in the time domain), but alas, it still contains some problems
that it's developers are still trying to work out. Not sure, but I don't
believe XTK has an equivalent t-line model.

So depending on what you are doing, one tool may be preferable over the other.
Or you may need both tools. I'm sure there are a lot of opinions on
the matter held by those on the list. The above is my take on the topic.

-Ray

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


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Wed Nov 22 2000 - 10:50:53 PST