Minutes, IBIS Quality Committee

14 Feb 2006

11-12 AM EST (8-9 AM PST)
Phone: 1.877.384.0543 or 1.800.743.7560
Meeting access: Meeting ID: 90437837

ROLL CALL
Adam Tambone
Barry Katz
Benny Lazer
Benjamin P Silva
*Bob Ross
Brian Arsenault
*Eckhard Lenski
Eric Brock
Gregory R Edlund
Hazem Hegazy
John Figueroa
John Angulo
Katja Koller
Kevin Fisher
*Kim Helliwell
Lance Wang
Lynne Green
*Mike LaBonte
*Moshiul Haque
Peter LaFlamme
Robert Haller
Roy Leventhal
Sherif Hammad
Todd Westerhoff
Tom Dagostino
Kazuyoshi Shoji
Sadahiro Nonoyama

Everyone in attendance marked by  *

NOTE: "AR" = Action Required.

-----------------------MINUTES ---------------------------
Mike LaBonte conducted the meeting.

- Review of minutes for 10 Jan 2006

Meeting date year listed as 2005, needs to be 2006.

AR: Mike update Jan 10 2006 minutes file.

- Announcements - EIA Call for Patent Disclosure: 

No company declared a patent .

- AR Review

AR: Mike write up straw man bug report and discussion for check 2.2

Done.

AR: Bob Ross contact Kim about prior meeting notes

Done. No notes available.

- Opens for new items

The IQ effort was not discussed at the recent DesignCon IBIS summit.
No status presentation was scheduled and the subject was not raised.
Hopefully there will be news to report at a future summit.

Feedback from Moshiul Haque, Micron Technology

Micron has used the IQ guidelines and has been adding IQ headers to files.
The checks are basically OK, but some requirements are too stringent.
Micron would like to achieve at least level 1 status for their IBIS files,
but some info is not not available to produce required IBIS data.
Specifically, dynamic overshoot pulse_time is not available. Bob pointed
out that any check can be waived with a comment of exaplanation. There
was discussion of the circumstnaces that make a waiver legitimate.

Questions and points raised include:
  o How many waivers do we allow?
  o Most files today don't have dynamic_overshoot.
  o Micron has never seen this request from a customer.
  o Which SI tools use dynamic_overshoot? SiAuditor does.
  o Dynamic overshoot info is found in some datasheets.
  o Could this check be marked optional?
  o IBIS is used for reflection, overshoot, SSN, and timing.
  o How about a waiver to accept models with static_overshoot only?

There is a significant gap in the difficulty of achieving level 1 status
for a file, compared to level 0. Additional intermediate levels might
alleviate this. For example:
  o A Acceptable file.
  o B Cosmetic blemishes that do not impact accuracy.
      Examples: Missing header text, unbelievable date, etc.
  o C Minor issues that can impact accuracy, use with caution.
      Examples: Missing pin RLC, missing or weak clamps, etc.
  o D File is mostly OK, with errors limited to a "damaged" section.
      Examples: Errors in a rarely used buffer, 1.8V models OK, 1.8V not.
  o E Errors found, file is broken.
Level D would not be acceptable from an IC vendor. Some discussion of
whether to use letters or numbers, and if the 1st code (A or 0) should
be for the best files or the worst.

Mike LaBonte asked why models with no power_clamp are prevalent. Points made:
  o Sounds like a bad extraction.
  o LCX technology really has no power clamp.
  o Some devices may be designed to tolerate high excursions, like 7V.
  o IQ check for clamp voltage range would catch this problem.
  o Bob Ross mentioned that he likes to remove non-monotonic portions
    of I/V curves.

There is an opportunity to update the IQ web pages so that they are more
similar to the Futures and Macromodel web pages. There appeared to be
little support for this at this time, since other items are of higher
priority.

Brief discussion of chairmanship. Barry Katz is listed officially. Mike
is acting chairman, but we need to identify a permanent chairman.

Next meeting:

Mar 7 2006
11-12 AM EST (8-9 AM PST)
Phone: 1.877.384.0543 or 1.800.743.7560
Meeting access: Meeting ID: 90437837

Meeting ended at 11:55 AM Eastern Time.