Minutes from the 25 April ibis-atm meeting are attached.
The following documents, which were discussed during the meeting, have been
posted as BIRD updates.
158.5 AMI Ts4file Analog Buffer Models
<http://ibis.org/birds/bird158.5.docx> Walter Katz, Signal Integrity
Software, Inc. Todd Westerhoff, Signal Integrity Software, Inc. Fangyi Rao,
Keysight Technologies, Inc. Radek Biernacki, Keysight Technologies,
Inc. February
20, 2013; May 15, 2013; May 17 2013; May 24, 2013; April 18, 2017; April
27, 2017
166.2 Resolving problems with Redriver Init Flow
<http://ibis.org/birds/bird166.2.docx> Walter Katz, Signal Integrity
Software, Inc. Darshan Shah, F5Networks, Inc. April 2, 2014; April 18,
2017; April 26, 2017
IBIS Macromodel Task Group
Meeting date: 25 April 2017
Members (asterisk for those attending):
ANSYS: * Dan Dvorscak
* Curtis Clark
Broadcom (Avago): Xingdong Dai
* Bob Miller
Cadence Design Systems: * Ambrish Varma
Brad Brim
Kumar Keshavan
Ken Willis
eASIC: David Banas
Marc Kowalski
Ericsson: Anders Ekholm
GlobalFoundries: Steve Parker
IBM Luis Armenta
Trevor Timpane
Intel: * Michael Mirmak
Keysight Technologies: * Fangyi Rao
* Radek Biernacki
Ming Yan
Maxim Integrated Products: Hassan Rafat
Mentor, A Siemens Business: 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
Kevin Li
Teraspeed Consulting Group: Scott McMorrow
Teraspeed Labs: * Bob Ross
TI: Alfred Chong
The meeting was led by Arpad Muranyi.
--------------------------------------------------------------------------------
Opens:
- None.
-------------
Review of ARs:
- Walter to send a BIRD 186.3 draft 1 to the ATM for review.
- Done.
- Walter to send BIRD 166.1 to Mike L. to be posted as a BIRD on the Open Forum
site.
- Done.
- Radek to send BIRD 158.4 (with the editorial changes from the meeting) to
Mike L. to be posted to the ATM archives as BIRD 158.4 draft 4 and to be
posted to the Open Forum site as BIRD 158.4.
- Done.
- Arpad to send an email to the Open Forum noting that ATM recommends that
BIRD 158.4 be accepted.
- Done.
--------------------------
Call for patent disclosure:
- None.
-------------------------
Review of Meeting Minutes:
- Arpad: Does anyone have any comments or corrections? [none]
- Radek: Motion to approve the minutes.
- Dan: Second.
- Arpad: Anyone opposed? [none]
-------------
New Discussion:
BIRD 158.4:
- Radek: A small correction was recommended at the Open Forum meeting.
- The example for Ts4File_Includes was not consistent with the keyword.
- I have a question for this group. Should we change the name of
Ts4file_Includes? Perhaps Ts4file_ExtendsTo?
- Curtis: How about Ts4File_Boundary?
- Radek: I like that suggestion.
- Bob R.: So you will change Ts4file_Includes to Ts4File_Boundary and correct
the example text?
- Radek: Yes and submit it as BIRD 158.5.
- Arpad/Bob R.: Can it be sent to Mike L. and posted to the Open Forum site by
Friday, so it will still be eligible for a vote at the next Open
Forum meeting?
- Radek: Yes, I will do that.
- Arpad: I'll remove this item from our ATM agenda.
BIRD 166.1 Redriver statistical flow:
- Walter: I understand the issue we have with crosstalk.
- In our current flow we don't handle crosstalk correctly when we have
repeaters. We certainly don't propagate it properly through redrivers, but
even for retimers we have issues.
- I agree. But I think this issue is independent of the non-crosstalk issue
we have with not getting the proper main channel IR to Rx2 Init().
- I think we need an additional IR out of each Tx Init() representing its
equalization, so the Tx can propagate crosstalk properly.
- I think they are independent issues.
- Fangyi: I want to correct one statement you made.
- In the current (IBIS 6.1) flow, crosstalk is not an issue.
- The issue with the current flow is that the final IR presented to Rx2 Init()
is wrong (it doesn't contain the upstream channel).
- But in the current flow crosstalk is okay. You can account for any
crosstalk path.
- In the current flow the input IR to Tx2 is the downstream channel.
- In BIRD 166.1, by changing the IR input to Tx Init(), you can't account for
all the crosstalk paths anymore.
- Discussion: Fangyi described an analogous bit-by-bit GetWave() flow in which
the same problem occurs: If you have a GetWave() redriver simulation and the
Tx2 model is Init() only. In that case, you need an IR that captures only the
downstream channel and the Tx2 equalization. This would be used to propagate
the output of Rx1's GetWave() to Rx2's GetWave(). If a flow like BIRD 166.1's
were used at Init() time, then the IR input to Tx2's Init() would contain more
than just the downstream channel, and the EDA tool would never have access to
an IR that captured only the Tx2 equalization and the downstream channel.
- Arpad: I thought this BIRD was only supposed to fix the statistical flow?
- Walter: My flow (at Init() time) would cause problems in the case Fangyi
described.
- Walter: I could change my BIRD (back to the original BIRD 166, in which the
the IR input to Tx2 Init() was still just the downstream channel).
- But then we are back to the problem if a Tx2 optimizes itself.
- Fangyi: In the future proposal the Rx returns its own equalization.
- If the Tx does the same thing, then we probably solve all the flow
combinations and can even eliminate the crosstalk IR matrix.
- Walter: I totally agree. I've been saying that the Tx should return its
equalization and the Rx can return its equalization in two parts
(LTI and non-LTI), and then the EDA tool can do everything.
- Arpad: With this change (rollback) we are now suggesting, the only problem
remaining will be the redriver Tx optimization case?
- Walter: Yes.
- Walter/Fangyi: The redriver Tx typically doesn't optimize itself anyway. The
only Tx likely to optimize itself is the terminal (initial) Tx.
- Arpad: Sounds like we are approaching a solution.
- Walter: I will get that update out today.
- Arpad: Next week we can review it here and hopefully send it to the Open Forum
with a recommendation to accept it.
- Bob R.: Does this close the problem "permanently", or will it interfere with
future work?
- Walter: We do this BIRD 166 now.
- Then we enhance it later with Fangyi's Rx modifications and my Tx
modifications to get the final solution.
- Arpad: BIRD 166.x is planned for IBIS 7.0.
- Fangyi's proposal will be needed on top of it at a later point.
- Fangyi/Walter: Yes.
- Arpad: In the agenda item for this topic we have two notes:
- "add equations"
- This refers to Fangyi's proposal.
- "consider rewriting the AMI 'front end' in the spec (flows)"
- Does this refer to BIRD 166 or Fangyi's proposal?
- Walter: The comment from some was that the AMI documentation in the spec
was hard to figure out.
- The suggestion was to rewrite the whole AMI introduction section.
- I think this would best be done with Fangyi's proposal.
BIRD 186.3 File Naming Rules:
- Michael M.: Discussion on the editorial and interconnect reflectors has been
about 189.2 and some of the rules it uses for file names.
- There are references to "filename" in a number of examples.
- Text was copied from EBD, package, etc., sections that are changed by
BIRD 186.3.
- A quick glance at BIRD 186.3 suggests that most of the concerns Bob R.
raised with BIRD 189 are covered by BIRD 186.
- There are still some consistency issues within BIRD 186 itself.
- Walter: When we did BIRD 186, Bob R. went through and found all the places in
the spec that used file names and needed to be made compatible with
BIRD 186.
- Perhaps Bob needs to do the same thing for 189?
- Bob R.: The issue discovered at the meeting was an issue with file names vs.
base names in the EBD and Package Model sections. Those would have to
be added to BIRD 186 and corresponding changes be made in BIRD 189.
- We are using conflicting terminology, and it has been bad since day one.
- Now that we are talking about it in general, we need to clean it up.
- Michael M.: Could we make BIRD 186 the location for making these tweaks, and
Bob and Walter, the coauthors, could work on it?
- If we do it in 186, then it becomes imperative to get 186 approved before
189.
- Walter: I think it could be editorial, but I can go either way.
- Bob R.: Better we do it in a BIRD and formalize it.
- Walter and I can handle the changes.
- We should standardize on the 186.3 draft 1 proposal as a starting point?
- Arpad: This involves changes to 186 and 189, and no new BIRDs, correct?
- Bob R.: Yes.
- Michael M.: To summarize, we wait for an updated BIRD 186.3 draft from Bob and
Walter.
- Mike L.: Motion to adjourn.
- Michael M.: Second.
- Arpad: Thank you all for joining.
AR: Radek to send BIRD 158.5 to Mike L. for posting by Friday April 28th.
AR: Walter to send BIRD 166.2 to Mike L. for posting.
AR: Bob Ross and Walter to create a BIRD 186.3 draft 2.
-------------
Next meeting: 02 May 2017 12:00pm PT
-------------
IBIS Interconnect SPICE Wish List:
1) Simulator directives