Minutes from IBIS Open Forum Meeting 3/17/95

Date: March 29, 1995

From: Will Hobbs (503) 264-4369, fax (503) 264-4210
         Will_Hobbs@ccm.jf.intel.com
         XTG Modeling Manager, Intel Corp., Chairperson, IBIS Open Forum
         Intel Corporation
         5200 NE Elam Young Pkwy, Hillsboro, OR 97124 USA
         and
         Derrick Duehren (503) 264-4299, fax (503) 264-4904
         Derrick_Duehren@ccm.jf.intel.com
         Intel Program Manager, IBIS Secretary

Subject: Minutes from IBIS Open Forum Meeting 3/17/95

Participants:
ARPA Randy Harr
AT&T Global Info Solutions Dave Moxley*
Anacad Steffen Rochel
Ansoft Henri Maramis
Atmel Corporation Dan Terry
Cadence Design Sandeep Khanna, C. Kumar*
Cadlab Ralf Bruning
Contec Dileep Divekar
Digital Equipment Corp. Barry Katz
EIA Patty Rusher
High Design Technology Michael Smith, Dr. Ing. Cosso
HP Palo Alto Tom Langdorf
HP EESof Karl Kachigan, Henry Wu
HyperLynx Kellee Crisafulli
IBM Jay Diepenbrock, Joseph Flanigan
IBM-Motorola alliance Lynn Warriner, John Burnett
INCASES Werner Rissiek, Olaf Rethmeier
Integrated Silicon Systems Eric Bracken
Intel Corporation Stephen Peters*, Don Telian, Will Hobbs*
                              Arpad Muranyi*, Derrick Duehren*, Tim Schreyer*
Interconnectix, Inc. Bob Ross*
Intergraph Ian Dodd, David Wiens, Walter Katz
IntuSoft Charles Hymowitz
Mentor Graphics Ravender Goyal, Greg Doyle
Meta-Software Mei Wong, You-Pang Wei, John Sliney
MicroSim Arthur Wong
National Semiconductor Syed Huq*, Raj Raghuraum, Atul Agarwal
NEC Hiroshi Matsumoto
North Carolina State U. Steve Lipa, Michael Steer
OptEM Engineering, Inc. Benny Leveille, Ken Ehn
Pacific Numerix Paul K. U. Wang
Quad Design Jon Powell*
Quantic Labs Mike Ventham
Racal-Redac John Berrie
Symmetry Martin Walker
Synopsys, Logic Modeling G. Bill Lattin
Texas Instruments Bob Ward*
Thomson-CSF/SCTF Jean Lebrun
UniCAD Canada Ltd. Stephen Lum
Zeelan Technology George Opsahl, Hiro Moriyasu

CC:
Intel Corporation Randy Wilhelm, Jerry Budelman,
                              Intel IBIS team

In the list above, attendees at the meeting are indicated by *.

Upcoming Meetings: The bridge numbers for future IBIS teleconferences are
listed below:
     Date Bridge Number Reservation #
     4/7/95 (916) 356-9999 461605
     4/28/95 (916) 356-9999 461609

All meetings are 8:00 AM to 10:00 AM Pacific Time (16:00 to 18:00 UTC). We try
to have agendas out 7 days before each open forum and meeting minutes out
within 7 days after. When you call into the meeting, ask for the IBIS Open
Forum hosted by Will Hobbs and give the reservation number.

NOTE: "AR" = Action Required.

------------------------------------------------------------------------------

Check-in, Intros, Announcements
There were no new participants.

There were no corrections made to last month's minutes.

Conference Management Systems transferred our remaining funds ($1,400) to our
EIA account.

Derrick told of a Cadence AWK script for generating IBIS-like model data from
SPICE runs.

AR Kumar -- Check into it. [Done] The script converts raw vi data of the 4
tables to Cadence's SigNoise format. Kumar sees no use for it in the IBIS
forum unless someone is willing to write a script to dump the data into an IBIS
file. In that case the process would be:

  Raw Data (already in IBIS type of column format)
  ----> script
        ----> MODEL statement in an .ibs file

New Agenda Items:
New keyword for differential I/O (Bob Ward)
Enhancements to the timing spec (Stephen Peters)
Diode Transit Times (Tim Schreyer)

Press Updates:
Design Superconference, one paper by Chipcomm mentioned IBIS throughout.
IEEE Transactions, February issue mentions IBIS once.

                        ***** NEWS FLASH! *****
        The March 16 EDN Magazine has a feature article on IBIS
        by Derrick, Will, Arpad, and Robin Rosenbaum.
                        ***********************

New Models Available:
National Semiconductor submitted more models to Michael for posting.

Progress toward enlisting new IC vendors
None reported.

Golden Parser 2.1 progress and release date
- IBIS filename case resolution
  DOS seems to be inconsistent with its treatment of filename case, especially
  when transferring it to other operating systems.

  Consensus is that the Golden parser should only submit a warning if
  filenames are not all lowercase. Kellee captured our consensus and read
  his notes on what we decided. All agreed. It is repeated below.

  Bob Ross pointed out that this discussion also applies to .pkg files and the
  .ibs (.IBS) extension.

AR Kellee -- Work with Paul Munsey to resolve and implement these changes.

****************** CASE-RELATED GOLDEN PARSER CORRECTIONS ********************

We agreed that some changes are needed to the new IBIS parser before final
release.

We also agreed that Paul should receive additional compensation for making
these changes.

The following changes are required in the new IBIS parser:
1) After a file name is read from the command line the file extension should be
tested to see if it is either .ibs or .IBS, if so attempt to open the file
using the name as typed on the command line. If this fails because a file with
this name is not present, convert the name and file extension to lower case and
attempt to open the file again. If this fails because a file with this name is
not present, convert the name and file extension to upper case and try again.
If this fails, then report file not found. If any one of the three attempts to
open the file succeed, then proceed to parse and check the file.

2) The program shall include a compiler switch to perform the following:

a) If the file opens, and the name typed in on the command line is not lower
case and the parser was compiled for anything other than a DOS file system
computer, than the parser shall reported a warning to the user indicating the
file name should be lower case. (how is that for a long sentence).

b) If the parser was compiled for a DOS file system computer than no warning
should occur.

c) The above also applies to any file names referenced from within the file for
example .pkg files in version 2.x IBIS. So three attempts to open the pkg file
should be made, first as entered in the file, then as lower case, then as upper
case.

d) File names specified within the .ibs file must be lower case. This includes
both the .ibs and .pkg file names. If they are not lower case a warning should
be issued indicating that file names should be lower case. (I believe we
already have this warning. Paul please verify.)

e) The parser should issue ONE warning if any tab characters are found in the
file. The Warning should inform the user that 'IBIS files should not contain
tab characters'.

>> Arpad/Derrick's subsequent recommended wording: "It is recommended that
>> IBIS files not contain TAB characters"

This is to prevent problems with editors expanding tabs differently and ending
up with files with lines over 80 characters because tabs were expanded with say
8 characters instead of say 3 as the author intended.

f) The parser should also read files from either DOS or UNIX environments
without error. This probably requires the files to be opened in binary mode
and a special 'gets' function to be written which recognizes any of the
following as an end of line character:
  0x0d
  0x0a
  0x0d followed by 0a
  0x0a followed by 0d
It was noted that files have been found which contain the following as an end
of line character: 0d0a for DOS, or 0a for UNIX and this causes the parser to
reject correct files.

7) The DOS end of file character 0x1a (control Z) should be ignored. This is a
problem for UNIX versions of the parser reading a file from a DOS environment
which was transported without stripping the end of file character.

8) If a tab is found in an IBIS file it should be treated as ONE 0x20(space)
character. This applies both to counting the total number of characters on the
line and treating it as a white space character. (I think the parser already
does this. Paul please verify).

***************************** END OF CORRECTIONS *****************************

- Tabs and Control-Ms in .ibs files
  Apparently files are modified during transfer to and from vhdl.org. Dave
  Moxley found some wrong characters (^M at end of line) that weren't in
  Arpad's original file. Arpad reviewed all the Intel files on vhdl.org and
  found all the files except two had Ctrl-Ms at the end of every line. One of
  these only had the ^M where the disclaimer was edited into the file,
  implying that the editing process was responsible. Further, the editing
  process can be done correctly as demonstrated by the one clean file.

  The extra ^M should not cause a problem according to the spec, but the 1.1
  parser has a problem with it.

  Bob Ward suggested changing the library procedure to include a ^M stripping
  process and add a rule to the next version of the spec to exclude ^M. Arpad
  and Bob Ross would like to see the Librarian run a script to strip any Ctrl-
  Ms on .ibs files before posting it to vhdl.org. ^Z is another concern,
  which can occur whenever files are transferred in binary mode.

  There are several ways to strip these out, such as a stock filter (arcr?) in
  UNIX, to get rid of the ^Ms. FTP will usually do the right thing when
  transferring between OSs.

  The recommendation is to ask Paul Munsey to change the Golden Parser read
  routines so they work with either DOS or UNIX, ignore ^Z, and accept CRLF in
  various combinations (0A, 0D are acceptable as end of line, 1A as end of
  file).

AR Kellee -- communicate this to Paul. (Instructions to Paul are in Kellee's
Case changes above.)

  For archive purposes, UNIX format will be used. Librarian should run a DOS-
  to-UNIX filter on all files such that all files on vhdl.org are UNIX
  compatible (w/o Ctrl-Ms and Ctrl-Zs).

- Tabs to spaces.
  Some editors convert tabs to spaces, violating the 80 character line length.
  We could put a flag in the GP to warns user that tabs are found and that
  they can cause problems with some systems that may convert these to spaces.
  "Your file may not parse correctly if..."

AR Kellee -- Inform Paul, Arpad will work with Paul to clarify. (Instructions
to Paul are in Kellee's Case changes above.)

  Need to recommend that tabs not be used in the next rev of the spec.

AR Arpad -- Create a BIRD to strongly discourage use of tabs. [Done, BIRD 26]

AR Michael Steer -- Convert all tabs to single spaces (and inform submitter
that these changes were made).

  Will suggested that we offer Paul an extra $500 payment for this incremental
  work. Kellee will talk to Paul to see if this is appropriate.

- Advanced package options
  Tabled for our next meeting.

- Other issues
  We agreed to target 4/15/95 to have the full release of the GP.

Status of EIA affiliation
  Pattie Rusher has sent out the "Join EIA IBIS" letter but it allows those
  who are already EIA members to escape paying dues. This was not our
  intention.

AR Derrick -- Talk to Pattie R. Correct the "If already member of EIA, no dues
due..." language in the EIA letter.

Rev 2.1 updates
o S2IBIS 2.1 Bob Ross reported that Michael Steer has a student assigned
              to work on the conversion utility, but the funding has not
              yet arrived. Bob Ross suggested that they modularize the
              program before expanding it further.

o Cookbook No discussion.

o Overview No discussion.

Model Creator Name
Will proposes that we recommended that the creator name or organization be
included as a comment in the .ibs file so users have a way of reporting bugs
found.

SGML/HTML Support?
PCIS (Pinnacle Component Interface Standard). Syed will post information on
IBIS/SGML after he collects more information.

New Keyword for Differential I/O
Bob Ward will submit an Egg.

Enhancements to the Timing Spec
Stephen will create an Egg.

Diode Transit Times
Tim Schreyer summarized his experiments at Intel with ESD clamp diodes and has
duplicated the effect in SPICE. The diodes overshoot when clamped, but have a
TT of up to 70 nS, and when they turn on, they stay on for 20 nS, then spike
charge back into the line when they turn off. This effect is very sensitive to
turn-off time. Tim has calculated and simulated with the same results.

Shottky diodes may exhibit different effects. Tim feels that Shottky diodes
won't exhibit this effect.

AR Tim -- "Clean up" your notes and give the files to Derrick for posting to
vhdl.org.

These method may be good to include as part of the IBIS spec. The forum will
continue its investigations, using Tim's paper as a starting point. We'll need
a way to measure it, and a way to get the information in IBIS, such as a new
keyword. TT, C0 and a few other parameters were used in the analysis. It
might be possible to measure this using a low impedance 4-foot line. Perhaps
we can generate TT from this measurement.

This could also be a good graduate study at NCSU.

Wrap-up, Next Meeting Plans
Our next meeting is a teleconference 4/7/95.
Suggested Topics:
o Directions for MCM.
o Golden Parser 2.1 progress and release date
  - Advanced package options
o New keyword for differential (Bob Ward's Egg)
o Enhancements to the timing spec (Stephen's Egg)
o EIA progress toward balloting

==============================================================================
                                      NOTES
If you know of someone new who wants to join the e-mail reflector
(ibis@vhdl.org), send e-mail to ibis-request@vhdl.org.

Check the pub/ibis directory on vhdl.org for more information on previous
discussions and results. You can get on via ftp anonymous, "guest" login from
telnet or dial-in (415-335-0110), or send an email request to the automatic
archive server, archive@vhdl.org.
==============================================================================
Received on Wed Mar 29 17:48:31 1995