Minutes, IBIS Quality Committee

27 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 change 4.1.4 to level 2
  Done

- David write 4.1.4a requiring 4.0 C_comp subparams to achieve level 4
  TBD

- Roy will send new version of his presentation to individuals for comment
  Done

New items:

Feedback on Roy's slides:
- The precision vs. accuracy slide is nice.
- Roy asks for any advice on new slides, particularly a correlation example.
- Mike corrected his statement from last week that he uses 5% of cycle time
  as the threshold for maximum allowable timing error. It is 1%.

Ongoing review of IQ checks:

2.2.	{LEVEL 2}  Do not use [Comment Char]
- This is not a quality issue.
- Moshiul has never needed to do this.
- Bob explained that one reason for [Comment Char] is that occasionally a
  pin name might use the pipe "|" character.
- Mike said that YACC-based parsers had trouble with this provision, but
  this should be resolved by now.
- We should either document the reason or take it out.
- We voted to strike this check.

2.3.	{LEVEL 2}  [File Name] is correct
- This is not a quality issue.
- IBISCHK catches this.
- This is really a level 1 check.
- Nevertheless, it is a common problem.
- This should be a level 1 check if we keep it.
- We could move this to section 5, which specifically discusses problems
  that are commonly seen.
- We voted to strike this, but add it to section 5.

2.4.	{LEVEL 2}  [File Rev] is correct
- This is not a quality issue.
- Mike thought we had overstepped our bounds in creating this check.
- Moshiul has been using the 0.x, 1.x, 2.x, 3.x convention, which is
  recommended by the IBIS spec.
- This scheme is redundant with IQ level modifiers.
- David suggested that this keyword, which does not prescribe a "hard"
  requirement, does not have the universal significance that Date does.
- We voted to strike this, but add it to section 5.

2.5.	{LEVEL 2}  [Date] is correct
- This is not a quality issue.
- It should be easy to get this one right, but hard for users to verify.
  - How does the user know it's wrong?
  - If the [Date] value in an IQ-certified file pre-dates the IQ spec,
    we know something is wrong.
- This is not a required IBIS keyword.
- We voted to strike this, but add it to section 5.

Mike suggested that section 5 should be made more prominent in the IQ spec.
- Mention it at the beginning of the document
- Put in an Appendix, easy for people to print as a reference.
- Helps model makers by pointing out the most common mistakes.

AR: David write 4.1.4a requiring 4.0 C_comp subparams to achieve level 4
AR: Mike make agreed 2.2 - 2.5 changes and post new draft spec

Next meeting:

05 Mar 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 12:11 PM Eastern Time.
