======================================================================
IBIS FUTURES/COOKBOOK TASK GROUP MEETING MINUTES
Date: January 5, 2006
http://www.eda.org/ibis/futures/
http://www.eda.org/ibis/cookbook/
Attendees:
----------
Cadence Design Systems - Lance Wang
Intel - Michael Mirmak, Arpad Muranyi
Mentor Graphics - John Angulo, Ian Dodd
Micron - Randy Wolff
Sigrity - Sam Chitwood
Teraspeed - Bob Ross
======================================================================
Next Meeting:
Thursday, January 12, 2006
9 AM - 10 AM US Pacific Time
Telephone Bridge Passcode
916-356-2663 1 325-1507
Agenda:
9 - 9:05 AM Opens
9:05 - 9:10 AM Macromodeling and measurements "short answer" (Muranyi)
9:10 - 9:20 AM BIRD100.1 Update (Muranyi)
9:20 - 9:30 AM BIRD102 Draft: File Name Extension (Mirmak)
9:30 - 9:55 AM Touchstone(R) Update Needs, Roadmap (Chitwood, Gupta)
========================================================================
Meetings have followed an irregular schedule due to recent US holidays.
The meeting covered a variety of short topics.
1) IBIS Cookbook update schedule: Lynne Green has described a variety
of issues causing confusion among Cookbook readers. Bob Ross suggested
that she write up a summary and a list of proposed changes. These could
be closed as part of the GEIA standardization/engineering bulletin
release process. Lynne to provide brief description of needed changes.
2) Expanding IBIS filename length per member requests: Michael mentioned
that at least two IBIS members have inquired about expanding IBIS filenames
in the same fashion that IBIS signal and model names have been expanded
(to approximately 40 characters). Bob suggested that, if these members
did not want to submit a BIRD, the Open Forum's regular participants could.
John observed that the numbers of columns in IBIS fields might be critical
to tool vendors. Arpad noted that, for IBIS creation tools, it makes sense
to extend filenames to match the modelname length. Michael to close issue.
3) Touchstone(r) specification closure: Sam Chitwood mentioned that
some small issues still remained with the Touchstone(R) posted specification;
Michael added that Stephen Peters had posted responses to public questions
on the IBIS reflectors, suggesting that version 12 of the draft document
was never approved. Team agreed to resume discussions with Agilent on
formally approving, if not standardizing, Touchstone(R). Michael to
contact Agilent and request participation. Sam may present on this to the
next IBIS Summit.
4) Multi-level Ethernet and need for measurements: Arpad Muranyi noted that
some inquiries were made on the reflectors regarding multi-level signaling
(e.g., Ethernet) and support under IBIS. This may involve the on-going
user-defined measurements enhancements to IBIS.
Ian Dodd stated that writing equations for both modeling and output makes
display options effectively endless. Arpad asked whether we have parameters
to describe these multi-level thresholds, etc.? Ian responded that we don't,
and for reliable operation, we need eye masks and multiple vinh/vinl parameters
at the least. Extending IBIS to do this is possible, but Ian suggests that
creating a measurement language to do this is a major effort.
The team asked whether we should reopen the user-defined measurement effort.
Ian noted that we have two approaches: with multi-lingual, the receiver can
determine when a transition has occurred. One reason for having IBIS-X was to
allow receivers to change the time at which they detect transitions based on
edge rate (ex. derating tables for DDR).
The second approach is to make measurements from the waveform. Bob suggested
making the mechanism call or user request definable in IBIS. Ian suggested
that we adopt an existing standard, such as GPIB, for reporting. Extending
IBIS would involve additional keywords, with the delays and perpetual
technology lag this implies.
Bob stated that we need to commit to a direction now, for the future. Ian
committed to an IBIS summit presentation on the topic, highlighting the need
to have measurements accompany models. The team asked whether the macromodeling
proposal from Arpad will support DDR2 derating easily. He will provide a a short
answer next time.
Michael noted that several other organizational issues are pending, including
expanding participation by current US-based members and options for overseas
participation. Should we hold multiple Futures meetings at different times or
perhaps regular "shifts" of meeting times? Bob expressed some doubt that this
would work, as continuity would be spoiled. Further discussion to take place
at an upcoming meeting.
Bob Ross also asked about DesignCon and DATE planning, plus financial
issues stemming from the Asian IBIS Summit. Time did not permit
discussion of these during the regular meeting.