===============================================================================

IBIS EDITORIAL TASK GROUP
Mailing list: ibis-editorial@freelists.org 

================================================================================

Attendees from September 6, 2024 Meeting (* means attended at least using audio)

ANSYS                                Curtis Clark, Juliano Mologni
Broadcom                             James Church
Intel Corp.                          Michael Mirmak*
                                     [Michael Brownell]
					Xiaoning Ye
Keysight Technologies                Ming Yan
Marvell                              Steve Parker
MathWorks                            Walter Katz
Micron Technology                    Justin Butterfield
Siemens EDA                          Weston Beal*, Arpad Muranyi*, Randy Wolff*
ST Microelectronics                  Aurora Sanna
Synopsys                             Ted Mido, Edna Moreno
University of Illinois               Jose Schutt-Aine
Zuken USA                            Lance Wang

Michael Mirmak called the meeting to order.  No patents were declared.

Michael provided an overview of the editing process, based on the IBIS 7.2 process.
The version checklist for BIRD 8.0 covers the known-issues list, all approved
BIRDs, any issues with those BIRDs, and tasks common to all versions, many
of which would be completed at the end of the process.

Michael noted that the SPIM and PSIJ BIRDs may require their own section.  An
open question is whether these BIRDs should be added at the end of the draft,
or before the older EMI section.

Weston Beal suggested that a separate set of keywords needs its own section.  
The organization should follow the PSIJ BIRD's defined keyword tree, similar to
the Interconnect Model Group.

Arpad Muranyi noted that Graham Kus preferred that IBIS versions maintain section
numbering consistent between versions, to ensure that tool documentation does not
become obsolete.

Weston suggested that the team could add SPIM and PSIJ after the EMI section without
problem.  Randy Wolff asked whether the team wanted a power integrity section for 
both BIRDs, not just a couple of new sections (including referencing existing keywords).

Weston replied that this was a good point.  This would be a Section 15 with power 
integrity parameters: Section 1 is PSIJ, Section 2 is SPIM, then add subsections (ISSO 
is one such keyword).

Michael asked which section should be placed first.  How should cross-referencing be
handled?  Michael accepted the AR to ask Kinger Cai.  [AR]

Arpad asked where BIRD 220 would go if approved: under Ch. 15 or [Model]?  Randy replied
that this would likely be under [Model].  Weston suggested a Power Integrity section would
need an introduction with cross-references, with PSIJ as sub-section 3 and SPIM as sub-section 2.

Arpad raised a concern about document stability due to its size and formatting, particularly
with change tracking.  Michael described an alternate approach to editing that supports
GitHub and the separation of content versus formatting.  This is being set up, but would
not be ready for IBIS 8.0 work.

Arpad noted that there were several non-technical changes raised in an August 23, 2024 
ATM/Editorial mail.  In the description of AMI_Init, one section describes the impulse 
matrix and aggressors (responses to crosstalk), while a second section talks about the Tx, 
and then the data passed into the Rx.  If you take impulse matrix as a whole, then the entire 
matrix from TX is passed into RX.

Weston noted that the size of the matrix in Tx and Rx may be different, but the key point is 
that step 2 contains the *modified victim*, which is returned from Tx to Rx.  Arpad replied
that the crosstalk stuff could be different data, and/or different in size.

Weston answered that the crosstalk is consumed by the Tx, but not returned; a different data 
set is returned by Tx AMI_Init; the EDA tool takes that in, combines it with crosstalk, then 
feeds this into the AMI_Init function for the Rx.  Michael suggested that a couple of new 
sentences would suffice (returned data is only the victim waveform).

Arpad noted that, in the second issue he discovered, the AMI_GetWave description starts well, 
but ends up only focusing on Rx.  Where is the TX AMI_GetWave process described?  He proposed 
having two paragraphs: one for TX, one for RX.

Michael asked where the TX disappears. The team discussed the paragraph featuring the text, 
"The sample values..."

Arpad added that there were also some issues in the text from a year ago related to +/- 0.5 V 
as the stimulus voltages; these may get changed due to jitter.  Weston noted that this was 
related to the definition of the stimulus as an analog sampled waveform.

Arpad stated that the output of the TX AMI_Getwave is a digital response convolved with the 
analog channel response (this is the change in IBIS 5.0 to 5.1; the waveform reporting location
moves from before Tx AMI_GetWave to *after* Tx AMI_GetWave).

Weston suggested that the output waveform *for the RX* is expected to be the waveform presented.
The first bullets should be:
* for Tx, it is this...
* for Rx, the following three conditions apply...

Weston suggested that the change doesn't require a BIRD but that the change should have multiple 
reviewers; a "kinda-BIRD" (Weston suggests this should be called an "absird"). Arpad added that 
the issue is related to jitter stimulus, which is also related to PAM and zero-crossing shifts.
Arpad took the AR to find the 0.5 V and zero-crossing statements in the IBIS 8.0 text and report 
the edits needed. [AR]

The team decided to meet weekly, except when the IBIS Open Forum teleconference is held.
Arpad moved to adjourn the meeting; Randy seconded the motion.  The meeting adjourned without objection.

The next meeting will be held September 20, 2024.