Minutes from the April 27, 2022 IBIS Interconnect Task Group meeting are
attached.
Regards,
Justin
================================================================================
IBIS INTERCONNECT TASK GROUP
http://www.ibis.org/interconnect_wip/ ;
Mailing list: ibis-interconnect@xxxxxxxxxxxxx
Archives at //www.freelists.org/archive/ibis-interconn/ ;
================================================================================
Attendees from April 27, 2022 Meeting (* means attended at least using audio)
ANSYS Curtis Clark*
Intel Corp. Michael Mirmak*
Michael Brownell
Keysight Technologies Ming Yan*
Marvell Steve Parker
MathWorks Walter Katz*
Micron Technology Justin Butterfield*
Randy Wolff*
Siemens EDA Arpad Muranyi*
ST Microelectronics Aurora Sanna
Teraspeed Labs Bob Ross*
University of Illinois Jose Schutt-Aine
Zuken USA Lance Wang*
Michael Mirmak convened the meeting. No patents were declared.
Justin Butterfield took minutes.
Ming Yan from Keysight introduced himself. He is a Research and Development
Engineer working on the IBIS model algorithms. He has recently worked on
implementing the IBIS-ISS in ADS.
Review of Minutes:
- Michael called for review of the minutes from the April 20, 2022 meeting.
Michael displayed the minutes. Curtis Clark motioned to approve the
minutes. Randy Wolff seconded. The minutes were approved without objection.
Review of ARs:
- Walter Katz to send out the updated port naming presentation.
- Done.
Opens:
- None.
Discussion:
PLS in Touchstone TSIRD Draft:
Arpad Muranyi added comments based on Walter's questions about swathing and
sparse matrix to the draft TSIRD. Arpad asked Vladimir Dmitriev-Zdorov about
these features, and he agreed these would be useful. Although, since the pole
residue format is much smaller in size, he was not sure on how much value
these features would add.
Arpad noted we need to decide where the pole/residue format will go in the
file structure. Michael asked Arpad to elaborate on the case where the
pole/residue and Touchstone data would be in the same file. Arpad noted that
the performance slow down can be mitigated by placing the pole/residue data at
the top of the file, as the EDA tool can stop reading the file at the end of
the pole/residue data. But, this does not solve the issue of the file size
and transmitting the models. Having all the data in one file has the
advantage of keeping the data together and reusing the port naming
information. We could go either direction or allow the model makers to do both.
Walter asked if we want to have the pole/residue data and Touchstone data in
the same file. Arpad replied that is what need to decide. Walter suggested
to have a format which has the pole/residue data and it can be optional to
include the Touchstone data.
Bob Ross commented that Touchstone data is the standard in the industry and
making it optional would make things more complicated. He suggested that
combining the formats can make sense only if the files are small. He
suggested to have two different formats for pole/residue and Touchstone.
Walter agreed that we want two different formats. He suggested we could have
a keyword in the pole/residue format that could point to the original
Touchstone file.
Randy asked if we would duplicate the port naming and other keywords that
overlap in the formats. Bob replied the keywords could be duplicated. Arpad
stated what he had in mind is that the existing keywords can be reused.
Michael asked if we can swap the Network Data keyword with the Pole/Residue
Data keyword. Walter suggested we should consider to just keep the data in
separate files. Arpad suggested the Touchstone data is useful for backwards
compatibility.
Bob noted the Touchstone data file could reference the pole/residue file.
Arpad suggested the file names could be the same between the Touchstone and
pole/residue. Bob noted we do not have consistency between .sNp and .ts for
Touchstone. Not all of the keywords would carry over to the pole/residue.
Bob asked if a draft could be sent out the reflector. Arpad was waiting for
some feedback from Vladimir. Arpad will send out the pole/residue format
draft [AR].
Arpad asked if we want a separate specification in IBIS for the pole/residue
format. Michael asked if we could get a list of keywords in Touchstone and
mark the keywords which are appropriate for pole/residue. Walter suggested we
could take the Touchstone 2 specification and delete all the unnecessary
information. Then, we can look at what is necessary for the pole/residue
format. Arpad will take a look at what will carry over from Touchstone 2 [AR].
Port Naming:
Walter suggested to keep the format simple and just support differential and
not Chord signaling. Arpad noted this seems similar to Pam4 vs. PAMN. He
asked why we cannot have differential with more than two nets. Michael asked
if there is demand for Chord. Arpad replied C-PHY is a common standard being
used in the industry.
Walter suggested we could have a group column. Arpad asked if we could remove
the plus and minus and only have a group name for the set. Walter stated, for
a group of three or more, you have to get the ports in the correct order.
Michael noted, for some differential interfaces, there is a need to know what
the plus and minus wires are.
Arpad noted for connecting EMD models all we need to know is if the port is a
BGA pin name or reference designator pin name. He did not see the need for
the differential information. Michael asked if we should make the
differential information optional. Walter agreed it could be optional, but we
should include a format for the differential information. Arpad asked if we
care about the differential. Walter noted the IEEE format has added the
differential information as comments.
Michael asked if Arpad could give an example of the physical connectivity
[AR].
Randy will send out links to the C-PHY information from previous IBIS Summits
[AR].
Randy said he can take over as chair of this meeting while Michael is out, but
he is not available next week. It was agreed upon to cancel next weekâs
meeting.
Bob motioned to adjourn. Arpad seconded.
Next Meeting:
The next meeting will be May 11.
================================================================================
Bin List:
1. Touchstone 3
2. Pole-residue support for Touchstone
3. Port naming