Minutes from the 20 October ibis-atm meeting are attached.
IBIS Macromodel Task Group
Meeting date: 20 October 2015
Members (asterisk for those attending):
ANSYS: * Dan Dvorscak
* Curtis Clark
Avago (LSI) Xingdong Dai
Bob Miller
Cadence Design Systems: * Ambrish Varma
Brad Brim
Kumar Keshavan
Ken Willis
eASIC David Banas
Marc Kowalski
Ericsson: Anders Ekholm
IBM Steve Parker
Intel: Michael Mirmak
Keysight Technologies: * Fangyi Rao
* Radek Biernacki
Maxim Integrated Products: Hassan Rafat
Mentor Graphics: 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
Teraspeed Consulting Group: Scott McMorrow
Teraspeed Labs: * Bob Ross
TI: Alfred Chong
(Note: Agilent has changed to Keysight)
The meeting was led by Arpad Muranyi.
------------------------------------------------------------------------
Opens:
- Arpad mentioned that the IBIS summit at EPEPs is next week (October 28), and
we might want to cancel next week's ATM meeting. Bob and Walter agreed, and
the group decided to cancel the ATM meeting on October 27th.
- Arpad noted that the new Special Parameters BIRD, BIRD 179, had been added to
the list of pending BIRDs at the bottom of the weekly ATM agenda.
--------------------------
Call for patent disclosure:
- None.
-------------
Review of ARs:
- Arpad to submit the final version of the Special Parameters BIRD to Mike L. to
be uploaded as BIRD 179.
- Done.
- Radek to start an email thread regarding finalizing the language for the
cleanup of references to "ground".
- Done.
- Walter to email draft 2 of the Redriver Init Flow proposal to Mike L. to be
posted to the ATM website.
- Done.
-------------------------
Review of Meeting Minutes:
- Arpad: Does anyone have any comments or corrections? [none]
- Todd: Motion to approve the minutes.
- Mike L: Second.
- Arpad: Anyone opposed? [none]
-------------
New Discussion:
Item #6: New Redriver Flow BIRD (untabled last week, was Item #11).
- Discussion: Walter and Fangyi mentioned that they had continued their
private discussions on the topic, and that they were happy with the progress
in understanding each other though plenty of work remained. Fangyi then posed
additional questions regarding Walter's proposal. He stated that the flow #1
defined in Walter's proposal should properly provide the combined IR to the
terminal Rx regardless of whether the redriver Tx self-optimized. He
therefore wondered if the newly proposed parameter was necessary at all.
Crosstalk contributions were then discussed. Fangyi stated that in the
general case analyzing a network should not restrict the locations at which
a given Tx could insert cross talk. Any Rx could receive crosstalk from any
Tx, regardless of their locations. A single Tx might be an aggressor on the
upstream Rx, which would then propagate down the channel to the terminal Rx,
and the same Tx might be an aggressor directly to the terminal Rx. Fangyi
felt that clarifications in the spec had stated that a Tx's IR matrix included
IRs from the Tx to all Rxs (regardless of where any Rx might be). Fangyi
restated his thought that the only outstanding issue was that the terminal Rx
does not receive the combined IR so that it can optimize properly, and that
this could be handled by passing in two IRs, the combined IR and the
downstream-only IR. He felt that this and all the crosstalk terms could be
handled without changing the footprint of the Init() call. Additional IRs
could simply be appended to the IR matrix passed into Init(). Walter said he
had an even better understanding of Fangyi's points, and the two could
continue to work on the issues.
Item #7: Language corrections regarding "ground".
- Discussion: Several people thanked Radek for composing an excellent email
summarizing the issues and providing a good starting point. Radek suggested
that we first discuss how we would proceed with the actual process of making
the changes. Mike L. reviewed his reply to Radek's original email and
proposed that we might draft a BIRD that referred to work-in-progress
revisions of the 6.1 spec as changes were made. He said that in the event
that 6.2 were to become an SAE standard, we would need a robust change
history. He noted that it was by no means certain that 6.2 would become an
SAE standard. He felt that in any event it might be tedious to track all the
changes for this project in our typical BIRD fashion. Radek stated that he
felt it might be best to follow the traditional BIRD process, and that Walter
had already provided a list of most, if not all, of the places that would
require revisions. Mike L. expressed concern that writing a traditional BIRD
would then rely on someone to laboriously fold all the described changes into
the spec. Walter suggested that we might edit 6.1 directly to create 6.2 and
then use the difference between that and 6.1 to generate the list of proposed
changes. Arpad agreed and said that the BIRD might then just list the page
numbers and paragraphs of the changes. Bob then concurred with Radek, and
noted that additional BIRDS like 179 would also be folded into 6.2. Bob also
noted that the pictures in IBIS 6.1 were done by manual writers hired by Tech
America, and that they were not internal figures that could be edited directly
in Word (they might be Visio). Mike L. suggested that we should get ahold of
the original figures and archive them if possible. Lastly, the group agreed
that work on this task should be handled within the ATM group meetings
themselves.
- Arpad: Now is a good stopping point.
- Thank you all for joining.
-------------
Next meeting: 03 November 2015 12:00pm PT
-------------
IBIS Interconnect SPICE Wish List:
1) Simulator directives
