[SI-LIST] : IBIS Accuracy Specification First Draft

gedlund@us.ibm.com
Thu, 18 Feb 1999 10:23:52 -0600

(OK, so nobody is interviewing us - but we thought some of you might be
interested in this "press release.")

First Draft of IBIS Accuracy Specification Unveiled at DesignCon99

February 1, 1999 - A group of engineers from system and semiconductor
companies has been working for the past year on a specification that would
allow an IBIS datasheet (model) provider to document the results of lab
correlation. This new effort is called the "IBIS Accuracy Specification,"
and we presented our first draft at the DesignCon99 IBIS Summit. The
documents are available on the IBIS web site at
www.eigroup.org/ibis/ibis.htm under the heading "Accuracy Info." The
intention of this specification is to offer the IBIS user a quantitative
method for assessing the accuracy of an IBIS datasheet. The user can then
make a data-driven decision about whether or not an IBIS datasheet and the
component it represents are appropriate for the design application. The
IBIS Accuracy Specification is a potentially powerful tool for specifying
the purchase of semiconductor components.

The basic idea is this: the IBIS modeling engineer measures a set of test
loads called out in the spec and compares the results against SPICE "golden
waveforms" that are embedded in the IBIS datasheet (model). The modeling
engineer then applies a metric that quantifies the comparison and places
the metric results in the IBIS datasheet as comments. The user can then
run a comparison against the same "golden waveforms" using the behavioral
simulator of his or her own choosing. So we have established a link
between behavioral simulation results and the actual hardware the IBIS
datasheet is supposed to represent.

We invite the IBIS and signal integrity communities to review the IBIS
Accuracy Specification and offer their comments and suggestions. Please
send them to ibis-users@eda.org. This document is a work-in-progress; we
are already working on revision 1.2. It is our intention to take the
specification before the Electronic Industries Alliance for approval as a
formal industry standard in 1999.

Furthermore, we would like to invite any of you who has an interest in IBIS
accuracy to join us in an open teleconference call beginning February 26,
1999. This call will be held the every third week one week after the
regular IBIS Open Forum call. A formal announcement will be posted to the
IBIS reflector after the February 19 IBIS Open Forum conference call. (You
may join the reflector by sending email to ibis-request@eda.org.) The
accuracy working group will address four main topics:

1) amending IBIS to include new golden waveforms
2) revising the IBIS Accuracy Specification to 1.2
3) pursuing EIA approval
4) expanding the scope to cover contemporary I/O buffer designs.

As IBIS users, you may also be interested in a test board that we designed
as a companion to the specification. It demonstrates test structures that
implement the measurements called out by the specification. You can
down-load the schematics, Gerbers, and Cadence design source files. The
board design is public domain, and we encourage you to build your own copy
or even change the design. Please read the errata first.

**** 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 ****