Minutes of: Verilog-AMS LRM Device Modeling Subcommittee Meeting May 6


Subject: Minutes of: Verilog-AMS LRM Device Modeling Subcommittee Meeting May 6
From: Geoffrey.Coram (Geoffrey.Coram@analog.com)
Date: Tue May 06 2003 - 11:46:38 PDT


Attendees:

Geoffrey Coram, Analog Devices
Ilya Yusim, Cadence
Jeroen Paasschens, Philips
Rick Poore, Agilent
John Moore, Agilent
Colin McAndrew, Motorola
Srikanth Chandrasekaran, Motorola
David Zweidinger, Texas Instruments
Marek Mierzwinski, Tiburon
Boris Troyanovsky, Tiburon
Patrick O'Halloran, Tiburon
Peter Liebmann, Xpedion
Han Koh, Xpedion
Mustafa Sungur, Nassda
Wolfgang Roethig, NEC
Gary Fedder, Carnegie-Mellon University
Ken Bakalar, Mentor Graphics
Wei-Dong Liu, Synopsis

1. Introductions

Attendees discussed their involvement with Verilog-AMS;
the two main camps were (a) those in device modeling groups
that do the hand-coding in C and (b) those in simulator
groups implementing Verilog-AMS for customer use.

2. Committee guidelines

The following items were listed as general guidelines for the
subcommittee:

- Recommend as little as possible, meaning that we prefer
  general solutions that address multiple needs.

- Eschew analysis- or simulator-specificity where possible.

Note that all recommendations will be sent to the full
Verilog-AMS technical committee (and then upwards within
Accellera) for final approval.

3. Goals

The attendees were asked, "What should an extension for device modeling do?"

  Geoffrey Coram: Extensions should help make a device model written in
  Verilog-AMS as efficient and fast as one coded natively in C/C++

  Ken Bakalar: Extensions should make implementations portable by
  using formal precision in specifying to the simulator what they
  mean/should do.

  Mustafa Sungur asked if the extensions should help speed up model
  development or speed up the run-time performance.

  Colin McAndrew said that by allowing re-use of modules, this will
  speed up the development, but the main goal of the subcommittee
  should be the run-time efficiency and removal of errors in
  implementation (perhaps due to ambiguous specifications, as
  per Ken's point)

  Wolfgang Roethig agreed that the extensions should facilitate
  writing portable, reusable models.

  Someone asked if we had contact with model developers; in some
  sense, the non-EDA people (TI, Analog Devices, Philips) are
  responsible for model development, but the point was more
  towards research groups writing new models.

  Colin listed several models at universities that are being
  written in Verilog-AMS, and said that the Compact Model Council
  (CMC) members had expressed interest.

  Marek Mierzwinski mentioned that the BSIM group at Berkeley is
  considering using Verilog-AMS, but the CMC wants models in C
  The extensions are mainly for improving speed, but also
  "convenience items" that the model developers have gotten
  accustomed to using.

4. Technology donations

   Colin is trying to get ADMS released into the public domain,
   either through "SourceForge" or this subcommittee.

   Someone pointed out that this committee is not really about
   developing software. However, the ADMS platform may be a
   useful example for us to evaluate how powerful a proposed
   extension is.

5. Digression into attributes

   Peter Liebmann suggested that we look into attributes in Verilog2000
   as a possible way to implement many of the proposed extensions.

   Someone pointed out that we really haven't generated a list of
   the extensions yet.

   Ken gave a quick tutorial for attributes, that they are
   sort of a magic comment that simulators may ignore (if they don't
   recognize) because they should not affect the results, only perhaps
   the efficiency.

   Wolfgang was concerned that attributes were poorly controlled
   and could introduce non-portability issues if they were
   implemented differently. Does "bias-independent" have a clear
   meaning, if it were used to label a named block to indicate
   that the expressions need not be recomputed each iteration?
   We need to codify some attributes to get portability.
   Also, he suggested that we look at other extensions that
   define semantics, such as in SystemVerilog.

   Ken metioned that there are other methods of syntax extensions
   besides attributes.
5. Suggested enhancements

  Srikanth Chandrasekaran suggested that we make up a list of
  problems we are trying to address.

  To that end, Geoffrey asked that anyone with a list of
  suggested enhancements e-mail them to him. Many contributions
  are already collated at
  http://www.designers-guide.com/private/vams-extensions/compact-modeling/index.
html

  Geoffrey hopes to generate a comprehensive "wish list" and then
  eliminate things that have insufficient justification for a
  language extension.

  (In an earlier context, Mustafa was urged to suggest extensions
  that would help a model run better in high-capacity simulators
  like Nassda's HSIM.)

  Wolfgang asked if the VHDL analog working group had done any work
  in this area, and if we could review it for applicability.

  Ken said he didn't think there was any such work. (If you know
  differently, please let Geoffrey know.)

6. Current implementations

  David Zweidinger asked about the memory efficiency for ADMS, Tiburon.
  What he's read has only addressed speed, which is slower but perhaps
  acceptably so. His company is already having memory problems with
  native C models for top-level sims.

  Boris Troyanovsky said he didn't think it would be an issue,
  because there were ways to improve the memory use; however,
  Colin and Boris admitted that they have not done formal
  benchmarking.

7. The next subcommittee meeting is set for Tuesday, May 20
   at the same time, 11 AM US-Eastern Daylight Time
   (8 AM US-Pacific Time)

  Wolfgang mentioned this was difficult for his company's
  modeling group in Japan; however, since they did not
  contact Geoffrey, he is reluctant to inconvenience those
  in Europe who did contact him. If there are people from
  Asia who are interested, we may push subsequent meetings
  to one hour later, 12 noon EDT (9 AM Pacific).

Srikanth asked that everyone in the call sign up for the
Verilog-AMS e-mail reflector. To subscribe, send e-mail
  To: Majordomo@eda.org
with the command in the body:
  subscribe verilog-ams your_address@company.com

The participants were all BCC'ed; if you got two copies
of these minutes, you're probably already on the reflector.

-- 
Geoffrey J. Coram, Ph.D.    Senior CAD Engineer     
Analog Devices, Inc.        Geoffrey.Coram@analog.com 
804 Woburn St., MS-422,     Tel (781) 937-1924
Wilmington, MA 01887        Fax (781) 937-1014



This archive was generated by hypermail 2b28 : Tue May 06 2003 - 11:54:51 PDT