Minutes, IBIS Quality Committee

30 Oct 2007

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

ROLL CALL
  Adam Tambone
  Barry Katz, SiSoft
  Benny Lazer
  Benjamin P Silva
  Bob Cox, Micron
* Bob Ross, Teraspeed Consulting Group
  Brian Arsenault
  David Banas, Xilinx
* Eckhard Lenski, Nokia Siemens Networks
  Eric Brock
  Gregory R Edlund
  Hazem Hegazy
  John Figueroa
  John Angulo, Mentor Graphics
  Katja Koller, Nokia Siemens Networks
  Kevin Fisher
  Kim Helliwell, LSI Logic
  Lance Wang, IOMethodology
  Lynne Green
* Mike LaBonte, Cisco Systems
  Mike Mayer, SiSoft
  Moshiul Haque, Micron Technology
  Peter LaFlamme
  Radovan Vuletic, Qimonda
  Robert Haller, Enterasys
  Roy Leventhal, Leventhal Design & Communications
  Sherif Hammad, Mentor Graphics
  Todd Westerhoff, SiSoft
  Tom Dagostino, Teraspeed Consulting Group
* Kazuyoshi Shoji, Hitachi
  Sadahiro Nonoyama

Everyone in attendance marked by *

NOTE: "AR" = Action Required.

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

Call for patent disclosure:

- No one declared a patent.

AR Review:

- none

New items:

- There will be no meeting next week since Mike is unavailable.

- Shohji asked to continue review of JEITA comments on IQ.
  - This covers checks that we have not looked at for some time, but we agreed.

Continued review of JEITA comments on IQ:

- Item #11: 5.3.3 {LEVEL 3}  I-V tables have reasonable numerical range
  - Shohji: JEITA members feel that 2 or 3 amps would be more reasonable
    than a 1 amp limit.
  - Bob: There is no good reason for this 1A limit
    - IBISCHK warns at 10 amps
  - Mike: The reason for the limit is that tools like the IBIS viewer have
    zoom problems.
    - With large currents it can be impossible to zoom in closely.
    - Bob: That is a software defect
  - Mike: Maybe this check is not needed
    - Bob: agreed
  - Moshiul: This is a max limit, right? (yes)
  - Mike: Extreme currents can also cause simulation inaccuracies
    - Some simulators convert tables to polynomial models such as B-spline.
    - Large spikes in data can cause DC errors in the polynomial.
- Item #12: 5.3.14. {LEVEL 3}  On-die termination modeling documented
  - Shohji: JEITA members do not want to document this.
  - Mike: It may be as simple as having "ODT" in the model name.
    - It is not always easy to detect ODT in datasheets.
  - Bob: Weak pullup/pulldown is usually identified by name
  - Moshiul: Sometimes ODT is in a submodel
    - Some tools don't support this well
  - Eckhard: Sometimes we see current and wonder if it is intended.
    - The purpose of this requirement is to confirm the intention
- Item #13: 5.3.17. {LEVEL 3} Correlate IV curves to combined curves.
  - Shohji: "S" & "M" are mixed up in the text
  - Agreed; this will be corrected.
- Item #14: 5.4.5. {LEVEL 3}  Output and IO buffers have should have 2 sets of V-T tables
  - Shohji: LVDS needs only 1 set of waveforms
  - Mike: A study by Arpad Muranyi shows that signal accuracy is reasonable
    for simulation circuits that mimic the waveform fixture well.
- Item #15: 5.5.9. {LEVEL 3}  [Ramp] dt is consistent with data sheet
  - Shohji: Data sheet does not give ramp dt
  - Agreed
  - Is there another datasheet value that might provide a good check?
- Item #16: 6.2. {LEVEL 3} C_comp checked in both input and output mode
  - Shohji: JEITA members do not want to disclose on and off C_comp
    - Reveals inner details of buffer structure.
  - This is more like a best practice recommendation.
  - Moshiul: We can model c_comp on/off using c_comp_pullup, etc.
- Item #17: 6.5. {LEVEL 3} Minimize waveform lead-in time
  - Shohji: Some simulators delete lead-in time, others have trouble
    - Minimizing lead-in time is a good idea.
    - "Hint ..." should be deleted.
    - The IBIS waveform test fixture may result in artificially slow edges.
  - Moshiul: There is still an overclocking issue.
- Item #18: 6.7. {LEVEL 3} Differential models contain appropriate waveform tables
  - Shohji: This is more like a comment
  - "There is a discussion ..." can be deleted.
- Item 19: 6.8.	{LEVEL 3} Models correspond to data sheet
  - Shohji: Datasheets often do not give model specifics.
  - Eckhard: If datasheet shows 3 model strengths and IBIS has only 1 model,
    something is wrong.
    - The idea is to detect when models are shared inappropriately among pins.
- Item 20: 7.0 Correlation levels
  - Shohji: Seems to come from IBIS accuracy handbook
    - May be too much for this document
  - Mike: We may be re-writing this
  - Shohji: JEITA members do not agree with FOM methods (7.1.5.3)
- Item 21: 8. Model Limitations and Model Maker Notes
  - Shohji: Is this just a comment?
  - Mike: Yes, it appears to be

Next meeting:

13 Nov 2007
11-12 AM EST (8-9 AM PST)
Phone: 1.877.384.0543 or 1.800.743.7560
Passcode: 90437837

Meeting ended at 12:07 PM Eastern Time.