Minutes, IBIS Quality Committee

22 Apr 2008

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

ROLL CALL
  Adam Tambone
* Anders Ekholm, Ericsson
  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
* Guan Tao, Huawei Technologies
  Gregory R Edlund
  Hazem Hegazy
  Huang Chunxing, Huawei Technologies
  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
  Randy Wolff, Micron Technology
  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:

AR: Mike contact Lance about golden waveform generation
- Done
  - Golden Waveforms are an IBIS 4 feature, and s2ibis is IBIS 3.
  - We could write an add-on program to s2ibis
    - This might use existing .spi files to create new .spi files
    - There may be an issue with simulator dependence
- An earlier proposal to use committee funds to develop new s2ibis was shot down
  - Mike: There should be no problem with outside development
- Mike proposed a new way to implement s2ibis
  - The input is an actual IBIS file
  - s2ibis parser only understands keywords needed to generate data
  - Some special "|s2ibis" comments would give simulator parameters
  - All other keywords are simply passed through literally
  - This would continue to work with future IBIS versions
- Bob: s2ibis should ibscheck the files
  - Would have to look at [IBIS Ver] to know which ibischeck to run
  - Mike: ibischk4 works with older files.

New items:

Mike: Should we post an IBIS file with Golden Waveforms?
- Bob: We should not promote any vendor
- Mike: How about a generic sample?
- Anders: Shouldn't we promote vendors who produce quality models?
  - There was disagreement on this
  - Mike: a labeling program like EnergyStar might be OK
- IQ G designation
  - Mike: We would need to exclude trivial load fixtures.
  - Bob: Pre-emphasis uses [Driver Schedule]
    - Only Golden Waveforms can show the total effect
    - Mike: [External Model] also benefits from golden waveforms
    - There is no special need for Golden Waveform on submodels
  - Mike: Like ISO9000, the process of producing golden waveforms should
    lead to quality improvements, even without requirements.
  - Bob: Would a file with 1 Golden Waveform pass?
    - Are diff Golden Waveforms required?
    - File could become large
  - Moshiul: We might presume that if one model correlates well the others
    in the file will too.
- Bob: Automation may be difficult
  - Sometimes need to run multiple data cycles to get stable results
  - Mike: Do Golden Waveforms start at T=0?
  - Anders: The IBIS specification says Golden Waveforms conform to
    Rising and Falling waveforms
  - Bob: Overclocked golden waveforms would be good
  - Mike: Should we file a BIRD for stimulus patterns?
  - Bob: Golden Waveform is overloaded already
- Bob: IBIS files are getting large
  - The Golden Waveform point limit is the same as for [Rising waveform] and
    [Falling waveform].
- Bob: It will take competing vendors to drive Golden Waveforms
  - This is a chicken and egg problem
  - Mike: Waveform display tools could read IBIS files and show waveforms.
- We will draft a G spec
  - Bob: Needs to require some minimum standard
  - For example, Golden Waveforms for only the serdes buffers
- Guan: How are Golden Waveforms used?
  - Bob: EDA tools have to simulate the same fixture using the IBIS buffer
    and compare to stored golden waveforms
- Moshiul: C_comp makes a difference at driver end
  - Mike: Agree
- Moshiul: The point count limit is a problem
  - Mike: We might want PRBS patterns
- Bob: maybe Golden Waveform data should be in external files
  - Anders: If we do this there will be more problems due to missing files.
- Anders: There is no option to correlate at the pin - pad only
  - Bob: The test load can replicate pin RLC
  - Mike: We might need as many test loads as there are signal pins.

AR: Mike write draft IQ G designation

Next meeting:

29 Apr 2008 11-12 AM EST (8-9 AM PST)

Meeting ended at 01:23 PM Eastern Time.