Minutes, IBIS Quality Committee

12 July 2011

11:00-12:00 EST (08:00-09:00 PST)

ROLL CALL

Cisco Systems:                * Mike LaBonte
Ericsson:                       Anders Ekholm
Green Streak Programs:          Lynne Green
Huawei Technologies:            Guan Tao
IBM:                            Bruce Archambeault
                              * Greg Edlund
IOMethodology:                  Lance Wang
Mentor Graphics:                John Angulo
Micron Technology:            * Moshiul Haque,
                                Randy Wolff
Nokia Siemens Networks:         Eckhard Lenski
Signal Consulting Group:        Tim Coyle
Teraspeed Consulting Group:   * Bob Ross
Texas Instruments:              Pavani Jella

Everyone in attendance marked by *

NOTE: "AR" = Action Required.

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

Call for opens and IBIS related patent disclosures:

- Bob: We will close a parser bug
  - Warning for missing model selector reference
  - It was already fixed on another bug report
  - The developer will document error codes

AR Review:

- Mike submit Test Keyword BIRD to open forum
  - Done

- Mike assemble multiple proposals for subckt call styles
  - Almost done

- Mike add a quail to correct open sink/source in IQ spec 5.6.2
  - Not done yet
  - Bob: This could be turned over to Mantis

- Bob conduct tests of parser behavior
  - No tests yet

New items:

AMI quality checks:
- Mike: We considered this for 2.0 but decided against it
- Greg: Getting a lot of inquiries about the state of AMI
  - Interacting with key suppliers
  - What do we ask them for?
  - There is no standard yet
  - Created a list of ideas

Greg shared "AMI Model Questionnaire":
- Greg: Item 5 is syntax checking
  - Item 13: Does the link train itself?
- Mike:
  - We try to avoid checks that IBISCHK checks
  - Also it should not just say "do what the spec says"
- Bob: This is a nice simple list
  - "Compliant" is an issue?
  - It may be too early to know what the model issues are
  - The spec has some problems
    - For example, examples inconsistent with spec
  - Case sensitivity rules may not be enforced by the parser
- Mike: This list might help find more parser checks
- Item by item:
  - 1 to 4
    - Bob: These are documentation
    - Greg changed 2 to "What company ..."
    - Greg colored items that could be on a checklist blue
  - 5
    - Mike: The parser can check this to some degree
      - It can't check model behavior compliance
  - 6
    - Mike: Isn't C_comp still applicable?
    - Greg: S-param models might have it
    - Bob: You have to know this before simulating
    - Mike: This might be a good check
    - Greg: Generic models don't have it
    - Mike: Good question for the ATM group
    - Moshiul: Isn't this like asking how many stages the driver has?
      - We don't do that
  - 7
    - Mike: Standard IBIS package models have to be used
    - Greg: Today you have to go outside IBIS
      - ISS would be good
  - 8
    - Greg: Sometimes the model comes from another company
      - I am supporting various IBM groups
      - Suppliers want requirements in the purchase agreement
  - 9
    - Greg added this
  - 10 to 13
    - Mike: More documentation questions
  - 14 to 16
    - Bob: The backchannel BIRD proposes this
    - Greg: This is about RX self training
    - Mike: Out params might be needed to report chosen settings
  - 17
    - This is about the hardware, not the model
    - Mike: Some vendors have model specific diagnostic outputs
  - 18 to 20
    - Mike: Standard golden waveforms should work for this
      - One question is if we need to go beyond time domain
    - Greg: We provide reports to our customers
      - Might be able to supply a cleaned up sample
  - 21 to 22
    - These are PHY specific questions
    - Mike: Are these hardware or models?
    - Greg: Hardware
- Mike: This could be split into datasheet and model groups
- Bob: A future question would be AMI version
- Greg inserted a new item 17: "DLL Checker"

AR: Greg send checklist to be posted

Mike: Have any AMI models not worked?
- Greg: Yes, I can ask people who have used them about this
  - Some models needed to be fixed up before they will run in other tools
  - We could mail the si-list asking for AMI model experiences
- Mike: We try to be careful about publicly probing bugs
  - The model review committee should be able check these
  - Not sure if Lynne has received

Next meetings:
- Next meetings July 19 and July 26

July 19 agenda:
- External Test

Meeting ended at 12:26 Eastern Time.