                                             
IBIS Macromodel Task Group

Meeting date: 21 Jun 2011

Members (asterisk for those attending):
Agilent:                    * Fangyi Rao
                            * Radek Biernacki
Ansoft:                       Chris Herrick
                              Danil Kirsanov
Ansys:                        Samuel Mertens
                            * Dan Dvorscak
                              Deepak Ramaswamy
                              Jianhua Gu
                            * Curtis Clark
Arrow Electronics:            Ian Dodd
Cadence Design Systems:       Terry Jernberg
                            * Ambrish Varma
Celsionix:                    Kellee Crisafulli
Cisco Systems:              * Mike LaBonte
                              Stephen Scearce
                              Ashwin Vasudevan
			      Syed Huq
Ericsson:                     Anders Ekholm
IBM:                          Greg Edlund
Intel:                        Michael Mirmak
LSI Logic:                    Wenyi Jin
Mentor Graphics:            * John Angulo
                              Vladimir Dmitriev-Zdorov
                              Zhen Mu
                            * Arpad Muranyi
Micron Technology:            Randy Wolff
NetLogic Microsystems:        Ryan Couts
Nokia-Siemens Networks:       Eckhard Lenski
Sigrity:                      Brad Brim
                            * Kumar Keshavan
                              Ken Willis
SiSoft:                     * Walter Katz
                              Mike Steinberger
                              Todd Westerhoff
			      Doug Burns
Snowbush IP:                  Marcus Van Ierssel
ST Micro:                     Syed Sadeghi
Teraspeed Consulting Group:   Scott McMorrow
                              Bob Ross
TI:                           Casey Morrison
                              Alfred Chong
Vitesse Semiconductor:        Eric Sweetman
Xilinx:                       Mustansir Fanaswalla

The meeting was lead by Arpad Muranyi

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

- None

--------------------------
Call for patent disclosure:

- None

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

- Arpad submit AMI_parameter BIRD to open forum
  - Done

- Arpad submit Bob's BIRD to open forum
  - Done

- Arpad add abbreviation periods to Corner Range BIRD draft
  - Done

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

Arpad showed the Format Corner and Range Clarification BIRD:
- Arpad: Are there any comments?
- Fangyi: Please clarify the 2nd paragraph on page 2
- Arpad: It instructs tools to use AMI corners
  - We may make changes later
- Radek: It would be better not to specify anything
  - It can be left to the EDA platform or the user
- Ambrish: There is more to this than it says
- Arpad: Then it would be ambiguous what typ, slow and fast mean
- Radek: The first 3 lines should be kept
  - The rest of the paragraph can go
- Mike: The last sentence can be kept
- Walter: The one before that can be kept too
- Arpad highlighted the middle sentences proposed for deletion
- John motioned to submit this with selected text removed
- Bob seconded
- The motion passed by acclamation

Arpad showed the task list:
- Arpad: This was updated an hour before the meeting
  - Everyone should check especially the cyan colored ones
- Arpad reviewed some of the tasks

- Item 35:
- Arpad: Should item 35 should in 5.1 or be pushed out?
- Walter: We will be more clear on this when when handle analog models
- Ambrish: Not sure how analog relates
- Arpad: There was talk about using Thevenin equivalents to generate impulse
  response
- Walter: The analog BIRD will satisfy people who want that
- Kumar: It could be a silicon model generating the impulse response
- Ambrish: AMI is supposed to be conjoined with IBIS
- Arpad: The spec currently says nothing about this
  - The question now is if this should be in 5.1
- Todd: A model without analog is non-compliant
- Arpad: Should that be said in the spec now?
- Ambrish: It says that now
  - Page 139
  - At most, an impulse response example might help
- Arpad: We will need graphics for that
- Ambrish: It could be ASCII
- Todd: The only fuzzy part of that is what the magnitude should be
  - This is more of a cookbook item
- Arpad: We can consider this to be tabled

Items 46 & 47:
- Walter: Some models use those
  - They have to stay
- Arpad: These will be marked "do nothing"

Item 56:
- Radek: Initialization is a good practice
- Arpad: I may not make sense to have it for Usage Out
- Bob: It is easier to leave it
- Fangyi: It doesn't hurt to have a default
- Arpad: We can put this off

Item 63:
- Walter: The AMI_Close function call is required
- Arpad: Disagree
- Dan: Close is required only if GetWave exists
- Arpad showed the IBIS 5.0 spec page 181
  "Case 3: Shared library has only AMI_Init"
- Walter: That should be deleted
- Todd: Agree
- Arpad: Should this be in 5.1?
- Bob: We should have a BIRD
- Todd: It should be solved soon
- Fangyi: It says elsewhere that AMI_Init is required

Walter showed his New Reserved Parameters for Data Management BIRD draft:
- Adds Supporting_Files, DLLPath, and DLLid
- Walter: Some models require supporting files
  - The EDA tool knows where the DLL file is, the DLL doesn't
- Ambrish: Is this Model_Specific?
- Walter: It is now
  - It needs to become reserved for wide usage
- Fangyi: This is a relative path?
- Walter: It could be absolute or relative to CWD
- Radek: This is not portable
- Walter: The Value is shown as NA
  - The tool will fill it in
- Fangyi: So it is not in the AMI file
- Walter: It is in the AMI file
  - The EDA tool replaces the value
- Fangyi: How does the tool know where the other files are?
- Walter: Files should be in the same directory as the IBIS file
- Radek: How do relative paths work?
- Walter: The DLL calls open() with whatever path it gets
- Kumar: Some vendors may not put them in the same directory
- Walter: IBIS requires it
- Ambrish: Agree
- Walter: DLLid is for multiple instances of the same DLL
  - The can be file writing collisions
  - This provides an ID to create unique names to avoid overwriting
- Kumar: For temporary files it has to use the ID?
- Walter: DLLs need to make sure they don't collide somehow
- Kumar: The tool could put the DLLs in different directories
- Radek: In what situation would this be needed?
- Walter: Where two models use the same DLL
  - The last parameter is Supporting_Files
  - This tells the tool to always move these files together
  - This can be used to support multiple versions of Matlab, for example
- Fangyi: Developers might be confused about what to use for DLLPath and
  DLLid.
- Arpad: How to give a different path for each DLL?
  - It doesn't say the path is relative
- Walter: Supporting files are relative to .ibs directory
- Arpad: Different OSes may be a problem
- Fangyi: Does DLLPath apply to supporting files?
- Radek: What is the Usage type for this?
- Walter: Info
  - Developers should check on all of this

Meeting ended.

-------------
Next meeting: 5 July 2011 12:00pm PT

Next agenda:
1) Task list item discussions

-------------
IBIS Interconnect SPICE Wish List:

1) Simulator directives
