Minutes, IBIS Quality Committee

06 feb 2007

11-12 AM EST (8-9 AM PST)

ROLL CALL
  Adam Tambone
  Barry Katz
  Benny Lazer
  Benjamin P Silva
* Bob Ross, Teraspeed Consulting Group
  Brian Arsenault
  David Banas, Xilinx
  Eckhard Lenski
  Eric Brock
  Gregory R Edlund
  Hazem Hegazy
  John Figueroa
  John Angulo
  Katja Koller
  Kevin Fisher
* Kim Helliwell, LSI Logic
  Lance Wang
  Lynne Green
* Mike LaBonte, Cisco
* Moshiul Haque, Micron Technology
  Peter LaFlamme
  Radovan Vuletic, Qimonda
  Robert Haller
* Roy Leventhal, Leventhal Design & Communications
  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.

AR Review:

- Mike update 2.1 check to clarify recommendation.

New items:

Logistics:
- We used MeetMe instead of MeetingPlace for this call:
  - This provides no application sharing.
  - The echo that plagued previous calls was absent.
  - We will use MeetMe again next time.

IBIS summit:
- Kim, Mike, and Bob attended.
- Thanks to Kim for presenting our group's process.
- After the presentation there were questions about:
  - correlation with simulation and measurements.
  - IQ checks for EBD, PKG, and ICM files.
  - Michael Mirmak sent email to Kim after the summit:
    - He likes the new syntax.
    - Information messages in IBISCHK are a good idea.
    - The IQ group should own the parser specification document.
- Bob gave his "X" presentation, giving examples of the interesting cases.
- There was some discussion of the IBIS-ATM effort.
- Moshiul asked if anyone from Nokia or STM was there.
  They have asked for IQ in IBIS files.

Parser specification document:
- This would be a good project for us after we wrap up the IQ spec.
- Bob reported that the CAUTION message is going in next software version.
- Kim's presentation mentioned the INFO flag
  - one pbug asked for CAUTION, the other for INFO.
- Bug93:
  - The parser rejects a correct model and accepts an incorrect one.
  - Submitted by Micron.
  - IBISCHK 4.2.1 will render the Micron IBIS library full of ERRORs
    - The ERRORs really are errors, not a mistake.
  - May make it a WARNING to accomodate those who are impacted.

Membership:
- Mike mentioned that Bob Haller may be joining calls soon.
- If there is interest we could have a new election for chairman.

ICM:
- We discussed why ICM adoption has been poor.
  - One theory is that the spec "got out of control" and became too complex.
- ICM might be used to give physical pin mapping for touchstone files.
- HSPICE has adopted at least part of the ICM syntax.
- Roy asked if the summit questioners have IBIS quality problems:
  - Michael Mirmak told Kim he does not know of such problems.
  - One ICM model from a connector company had errors.

Correlation:
- Roy has prepared slides on verification methods:
  - Survey of methods: FSV, etc.
  - Will not be posted because it is targeted for a specific conference.
- Need to continue parallel work in this area.

AR: Roy will contact David to continue collaborating on correlation.

IQ check 2.1 Discussion:
- Mike proposed that unavoidable IBISCHK WARNINGs should not require
  the X designator, but all others should:
  - Moshiul felt that this policy would be helpful.
  - Most felt that IBISCHK WARNINGs always need the X designator.
- Bob sometimes just cleans up clamp curves without comment.
- We discussed the merits of commenting out IBIS keywords that cause
  trouble vs. leaving the file with errors.
- From the model maker's point of view there are 3 possible actions
  when WARNINGs appear:
  - 1) Warnings can be avoided by cleaning up, X if you choose not to.
  - 2) For unavoidable warnings the file can be left broken, with an X.
  - 3) The offending parts could be commented out to avoid unavoidable
       warnings, with an X.
    - In this case some other check would no doubt be violated, since
      the file would be missing needed information.

AR: Bob and Moshiul will collaborate to rewrite section 2.1

Next meeting:

13 Feb 2007
11-12 AM EST (8-9 AM PST)
We will use Meetme again
Phone: 1.877.384.0543 or 1.800.743.7560
Passcode: 90437837

Meeting ended at 12:21 PM Eastern Time.