================================================================================ IBIS INTERCONNECT TASK GROUP Mailing list: ibis-interconnect@freelists.org ================================================================================ Attendees from September 20, 2023 Meeting (* means attended at least using audio) ANSYS Curtis Clark Broadcom James Church Intel Corp. Michael Mirmak* Michael Brownell Keysight Technologies Ming Yan Marvell Steve Parker MathWorks Walter Katz* Micron Technology Justin Butterfield Siemens EDA Arpad Muranyi*, Randy Wolff* ST Microelectronics Aurora Sanna Teraspeed Labs Bob Ross* University of Illinois Jose Schutt-Aine Zuken USA Lance Wang* Michael Mirmak called meeting to order and took minutes. No patents were declared. Arpad Muranyi asked about the recent Jose Schutt-Aine presentation and its treatment of normalization. Michael clarified that the presentation S-parameters to normalized Z-parameters. Bob Ross noted that S-parameters are always normalized in all Touchstone versions. Bob also noted that Michael was correct in the last meeting to assume unnormalized and normalized-to-1 network parameters are identical. Bob earlier had not realized that the port entries can be converted via a simple equation rather than matrix. Michael reviewed the minutes with the attendees. Arpad moved to approve the minutes. Bob seconded. No objections were raised and the minutes were approved. During Opens, Arpad noted that he had more details on port mapping which he will share later in the meeting. During AR review, Michael noted that his two ARs were still open: to contact Sam Chitwood on port referencing and loop inductances, and to contact Pete Pupulaikis on S-parameter compression algorithms. No other ARs were captured. Michael raised an open suggested by Arpad offline: how should the TSIRDs recently approved by the Open Forum be handled? Should Touchstone 2.1 activities start immediately? Arpad added that he recalled a one-slide Touchstone motion that was previously circulated. Should this be raised for voting at the next Open Forum? Michael showed the one-slide document. Randy Wolff noted that motion had been sent out by Lance Wang July 10 and was discussed in the Open Forum July 14. The minutes show this was brought up and shared but no action was taken, nor was an official motion or vote made. Walter Katz asked what bad things would happen if we wait to work on Touchstone 2.1. The document includes no other changes aside from referencing. Arpad observed that he had talked to his developers. Their treatment of referencing was similar to that of Synopsys HSPICE, with "R" followed by one or more numbers. Their format uses the committee parser for this. It would be preferable to get an updated parser in the short-term with the impedance improvements. Walter voiced his support for this. Randy suggested updating the slide and providing it for the next Open Forum. [AR] Lance to re-send the motion to the reflector for the next Open Forum meeting. Bob asked when Editorial meetings should start. Michael suggested alternate weeks with Interconnect, as Fridays are somewhat difficult. Walter asked what the starting point for the document should be. Michael wondered whether the Touchstone 1.x & 2.0 split document was appropriate. He also asked about reformatting the keyword descriptions to match the style used for IBIS. Arpad replied that he preferred the team start with Touchstone 2.0, without splitting between 1.x and 2.0. Randy added that TSIRD 3 and 4 were already present in a Touchstone draft. Walter proposed merging these two new TSIRDs into existing draft 10 for TSIRD 3 and 4. Who will do this work? Arpad noted that he was spending a lot of time on pole-residue updates. Michael volunteered to edit TSIRD 5.1 and 6.0 into draft 10 for a Touchstone 2.1 [AR]. Arpad shared some additional port-mapping proposal information. To clarify the distinction between flexible and fixed port usage: - if how a device is used is unclear, this is "flexible" usage; this approach may involve reserved names, however - if package models, MCM files, DIMMs, etc. are involved: this is "fixed" usage; ports use complete connectivity information Arpad added that the difference between "side" and port descriptions is not clear. This can be discussed. Arpad showed a proposed syntax, with masking of all but critical features and differences. - CONNECT covers extended nets as-is. - DIFFPORTS could be added; this consists of lists of port numbers with : and ; characters to identify port associations. - PORT DESCRIPTION has been added - SIDE has been added The PORT keyword is already present, and many examples were rewritten. Arpad has added transistor and differential clock examples. Remaining issues include: EMD was used as examples, but not IBIS Interconnect, and large, significant differences exist between IBIS Interconnect and EMD approaches. IBIS uses pin, pad, buffer, to cover package and on-die interconnects (three interfaces). No refdes concept exists in IBIS. IBIS connections are made by pin names and pad names, etc., some of which don't exist in EMD. EMD has only two interfaces between buffer and pin. Arpad asked the participants to review what is shown. Blue highlights in the document show what is not present in existing proprietary syntax. An example shows S-parameter in an IBIS package model using blue text. Walter noted that humans can read the net names, but an AI could do so as well. Arpad observed that he is not yet finished. He will match up the current syntax with the IBIS Interconnect syntax and intent. Walter asked about current users of Touchstone. Arpad replied that they can use refdes-based syntax. Walter thinks of packages as using three refdes declarations, for buffer, die pad, and pin. Arpad asked about the flow of modeling events He had in mind that Touchstone is generated first, then users want to hook it up to IBIS, for instance. Walter replied that a human would use names along with "Side" indicator to match IBIS to the Touchstone data. The human user will want a name for use on a schematic. Arpad commented on the Side concept: if we need Side, we could put that into the "Side" entry at the top. Arpad moved to adjourn; both Bob and Walter seconded. The meeting adjourned. The next meeting will be held Wednesday, September 27, 2023. ================================================================================ Bin List: 1) [Complete draft Touchstone document separating version 1.0 and 2.0 into their own chapters] - REMOVED 2) Create structures to encapsulate Touchstone 1.0 data in Touchstone 2+ specifications - TABLED 3) Complete draft Touchstone 2.0 document containing TSIRD3 and TSIRD4 draft (Muranyi) – COMPLETED IN DRAFT 10 4) Complete pole-residue format BIRD draft (Muranyi) 5) Complete port naming proposal (Katz) 6) Create alternatives to the Touchstone 1.0 option line before the "R" character - TABLED 7) Complete ISS-IRD 1 Draft - Enable Cascading of S-parameters Through W-element (Mirmak) - TABLED 8) Complete/revise Touchstone 3.0 draft outline (Mirmak) – dependent on several items above Tabled ARs: - Arpad to give an example of the physical connectivity needed for EMD automation.