IBIS Macromodel Task Group

Meeting date: 30 June 2015

Members (asterisk for those attending):
ANSYS:                      * Dan Dvorscak
                            * Curtis Clark
Avago (LSI)                   Xingdong Dai
                            * Bob Miller
Cadence Design Systems:     * Ambrish Varma
                              Brad Brim
                              Kumar Keshavan
                              Ken Willis
eASIC                       * David Banas
                              Marc Kowalski
Ericsson:                     Anders Ekholm
IBM                           Steve Parker
Intel:                        Michael Mirmak
Keysight Technologies:      * Fangyi Rao
                              Radek Biernacki
                              Nicholas Tzou
Maxim Integrated Products:    Hassan Rafat
Mentor Graphics:            * John Angulo
                            * Arpad Muranyi
Micron Technology:          * Randy Wolff
                              Justin Butterfield
QLogic Corp.                  James Zhou
                              Andy Joy
SiSoft:                     * Walter Katz
                              Todd Westerhoff
                            * Mike LaBonte
Synopsys                      Rita Horner
Teraspeed Consulting Group:   Scott McMorrow
Teraspeed Labs:               Bob Ross
TI:                           Alfred Chong

(Note: Agilent has changed to Keysight)

The meeting was led by Arpad Muranyi.

------------------------------------------------------------------------
Opens:

- None
--------------------------
Call for patent disclosure:

- None

-------------
Review of ARs:

- Mike LaBonte to post Walter's draft GND document.
  - Done.
  
- Mike LaBonte to send his BIRD spreadsheet to Arpad.
  - Done.
   
- Arpad to add review of minutes to the weekly ATM agenda.
  - Done.
  
- Arpad to add all BIRDs from the spreadsheet to the ATM agenda's tabled list.
  - Done. [See first new discussion item below]
  
- Arpad to contact Romi Mayder regarding BIRD 157.
  - Done.  Received out of office bounce back.
  
-------------------------
Review of Meeting Minutes:

- Arpad: This is the new agenda item we agreed upon last week.
- Curtis: I received no comments or corrections for last week's minutes.
- Arpad: Does anyone have any comments or corrections? [none]
- Mike L: Motion to approve.
- Arpad: Second.  Anyone opposed? [none]

-------------
New Discussion:

- Arpad: [showing the newly added BIRD list at the bottom of the agenda]
  - Mike L.'s spreadsheet had more information for each BIRD.
  - Columns for active, inactive, tabled, discussed, etc.
  - Was I supposed to keep all that, or is this list enough?
- Walter: People can use the website to find additional info.
- Mike L.: I think the goal was just to make sure they're all listed somewhere.
  - I'm okay with just having a list.
- Walter: I think we can leave it to the discretion of the chair.
- Arpad: We will leave it as is for now.

- Arpad: We were planning to talk about LTI analog models for AMI.
  - Scott informed me that he couldn't make it today.
  - We can discuss this in the future.
  
GND cleanup in IBIS:
- Arpad: Walter has a presentation.
- Walter: [sharing "GND BIRD" presentation]
  - Expect an IBIS 6.2 that is dedicated to cleaning up "ground" issues in 6.1.
  - Issues so endemic it's worth a separate task.
  - We should agree on the charter for this task for the editorial committee.
  - IBIS currently makes implicit assumptions about "Ground."
    - References to GND, Ground, Reference Node, Node 0, A_gnd, absolute ground,
      all need careful review and documentation.
  - When IBIS was originally written, Ground was often interpreted as truly
    global.
    - Confusing.  Is it node 0?  How are things connected to Ground?
    - In IBIS 1.0, C_comp is defined as "silicon die capacitance."
    - IBIS 1.0 doesn't say anything about how C_comp is hooked up.
    - My interpretation from 1.0 is that "Ground" was interpreted as node 0.
  - GND needs to be reviewed:
    - Under GENERAL SYNTAX RULES AND GUIDELINES, it says:
      ... following words are reserved and "must not be used"...
      - This list includes "GND".
      - This is not what we meant.  We meant "GND" is a reserved [Model] name.
    - GND is often used as a signal_name in [Pin] examples.
      - In this case it is the data book name.
  - IV Tables Reference Rail Voltages in Simulation.
    - The voltage used is between the signal node and the appropriate power rail
      buffer model terminals.
    - This should not be confused with derivation methods.
      - GND, Ground, Absolute Ground, or static voltages referenced to Test
        Fixture Ground.
    - GND, the ground symbol, and keywords such as [POWER_Clamp_Reference] are
      voltages relative to the Test Fixture Ground.
  - Simulator Netlists:
    - The intent is to make sure IBIS models are only looking at voltage
      differences between their terminals' nodes.
    - Don't let node 0 enter into it.
  - Simulator Reference Node:
    - Simulators usually have an absolute reference node (node 0, etc.).
    - I/O buffer or interconnect models should not use this node, and certainly
      should not supply or draw current from it.
    - This is the fundamental issue we've been discussing.
  - What is C_comp connected to?
    - IBIS 1.0 said nothing about it.
    - In IBIS 6.0 we find many locations:
      - GND, GND symbol, Text Fixture GND, GND_Clamp_Reference, etc.
    - There has been an implication that simulators should connect C_comp
      to the simulator reference node.
    - This implication is incorrect.  It's an interpretation based on the text
      used to describe derivation of the IV curves.
    - IBIS 6.2 should say C_comp should be connected to the local buffer ground
      rail.
  - I'd like to have everyone review it and understand it.
- Arpad: I'm in complete agreement with the final point.
  - C_comp should be connected to the buffer's local ground rail.
  - When I first saw simulators' implementations connecting C_comp to node 0, I
    felt this was not what we intended.
- Walter: We have representatives of most of the simulator companies here.
  - We could take the following interpretation:
    - The original intent for C_comp in IBIS 1.0 was capacitance at the die, and
      a reasonable interpretation is that it should be connected to a local
      ground rail.
  - There have been some strong statements to the contrary, but the simulator
    companies should all come to agreement on this issue.
  - There was no understanding of these requirements for IBIS 1.0 and it wasn't
    made clear.  Now is the time to make it clear.
  - Randy and I discovered that you get excellent correlation with SPICE models
    if C_comp is connected to the rail terminal.  This is not true if C_comp is
    connected to node 0.
  - Compelling technical reasons that it's the right thing to do.
  - We can justify that it was always the right thing to do.
  - This is the only controversial part of the presentation document.
  - Next step is further discussion here after conferring with our respective
    companies.
- Arpad: I would like to suggest a clear AR for all the EDA companies.
  - Check internally to see if we agree C_comp should not connect to node 0.
- Walter: Could someone make sure representatives of Synopsis and Zuken get the
          question?
- Arpad: I'll take the AR to send an email to the reflector.
  - Walter, would you forward it and respond to them privately?
- Walter: Yes.
- Arpad: Any other comments on this topic? [none]

- Arpad: Any motions to untable other topics to discuss? [none]
- Arpad: Let us close today's meeting.
  - Thank you all for joining.
  
AR: Walter to send his "GND BIRD" presentation to the reflector.
AR: Mike LaBonte to post the "GND BIRD" presentation.
AR: Arpad to send an email to the reflector regarding EDA vendor input on C_comp
    not being connected to node 0.
AR: Walter to forward/reply Arpad's email privately to Zuken and Synopsis to
    ensure they get the request for input.

-------------
Next meeting: 7 July 2015 12:00pm PT
-------------

IBIS Interconnect SPICE Wish List:

1) Simulator directives
