Re: [IBIS] What is IBIS ?

Mithlesh,

Most people who use IBIS have some familiarity with SPICE, so I'll give an
answer from the point of view of a SPICE user.

Non-encrypted SPICE models reveal a lot about one's IC design; not only the
circuit design (what transistors are connected to where), but some hints
about the technology of their Fab line. IC vendors don't like giving that
kind of information away. Some vendors use encrypted models, but few
simulator programs have that capability. (I know of only one.)

The IBIS equivalent of a SPICE model is an IBIS model, only we don't like to
call it an IBIS model. Actually the model is within the (IBIS-aware)
simulator, and the IBIS file that the vendor provides of his device, is a
special kind of data sheet that is formatted in a way that the IBIS model in
the simulator understands. The IBIS Spec. describes the format of the IBIS
data sheet.

IBIS was supposed to make it possible for people to freely give out IBIS
data sheets without fear of giving away trade secrets. However, some
vendors require that you sign an NDA before giving you access to their IBIS
models ... which kind of defeats the purpose of IBIS, as far as I am
concerned.

Rather than providing structural and connectivity information, as a SPICE
model does, IBIS data sheets provide only behavioral information. The basic
ground rule, upon which IBIS is based, is that the most fundamental
characteristics of an I/O or output buffer can be described by static I-V
curves, that describe the output buffer's drive strength or an input
buffer's clamping characteristics. If you have a device's I-V curves, you
have perhaps 90% of the necessary information about that device, and you can
predict reasonably accurately how that device behaves under transient
conditions. You need a little more information too, such as how quickly it
changes from the low state to the high state and vice-versa. That's where
the IBIS Ramp and V-T tables come in. Plus some information about the IC
package's electrical characteristics. Given all this, you can predict the
waveforms with arbitrary loads.

IBIS was designed for interconnects between ICs, so it models only the
output pin or input pin (that is, the characteristics of an IC as seen from
the outside of the I/O pins), but none of the core electronics like SPICE is
capable of doing.

In a simulator, then, you can connect IBIS models to interconnect (board
trace or wire) models and other device models, and simulate the whole thing.
And it probably simulates faster than it would with SPICE structural models.

Unlike SPICE where the SPICE simulator came first, and then came SPICE
models, there is no "IBIS simulator." IBIS is only a way for describing the
"models." Anyone who makes a circuit simulator can enhance it to make it
accept IBIS "models."

Because there isn't a standard for how an IBIS "model" should be simulated,
there can be some difference in results from one simulator to another.

IBIS data sheets are easily created from measurements of actual devices.
Doing this is often preferable to converting from a SPICE model to an IBIS
data sheet.

Also, it can be a lot easier to create an IBIS data sheet of a random
device, than it is to create a SPICE model for that device.

There are a lot of bad IBIS models out there. It's true that there are some
bad SPICE models out there too, but not as many. There is a lot of
ignorance about creating decent IBIS models.

Regards,
Andy

-----------------------------------------------------------------
|For help or to subscribe/unsubscribe, email majordomo@eda.org
|with the appropriate command message(s) in the body:
|
| help
| subscribe ibis <optional e-mail address, if different>
| subscribe ibis-users <optional e-mail address, if different>
| unsubscribe ibis <optional e-mail address, if different>
| unsubscribe ibis-users <optional e-mail address, if different>
|
|or email a request to ibis-request@eda.org.
|
|IBIS reflector archives exist under:
|
| http://www.eda.org/pub/ibis/email_archive/ Recent
| http://www.eda.org/pub/ibis/users_archive/ Recent
| http://www.eda.org/pub/ibis/email/ E-mail since 1993
Received on Wed Dec 22 07:26:14 2004