================================================================================ IBIS INTERCONNECT TASK GROUP Mailing list: ibis-interconnect@freelists.org ================================================================================ Attendees from March 1, 2023 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. Review of Minutes: - Michael called for review of the minutes from the February 22, 2023 meeting. Michael displayed the minutes. Arpad Muranyi moved to approve the minutes. Lance Wang seconded. The minutes were approved without objection. Review of ARs: - Arpad to give an example of the physical connectivity needed for EMD automation. - Arpad suggested to table this AR. Michael agreed we could create a tabled AR list. Arpad moved to table this AR. Walter Katz seconded. There were no objections. - Michael to check with Synopsys on permission for using the W-element S-parameter feature and what they can share on the algorithm details and model restrictions. - Michael reported he had a discussion with Ted Mido. Synopsys is not able to share details about the algorithms. Michael had asked if there were additional requirements for the W-element S-parameter data. Ted recommended a point at DC and low frequency points up to the point where the phase starts changing. Synopsys does not require that the W-element length be an integer. Arpad asked if Synopsys would be okay to allow usage of the syntax in IBIS-ISS. Michael replied that they are okay with it. Ted did request the addition of a node map, which maps the input and output nodes of the S-parameter to the W-element. Michael noted we would need to be careful with the port mapping in the S-parameter which could be overwritten by the W-element. Michael suggested to close this AR. - Michael to check if the Network Analyzer manuals documenting the original Touchstone can be posted to the IBIS website [AR]. - Michael noted the Keysight legal team is checking on this. He suggested to keep this open. - Everyone to think about what assurances we have that Touchstone X will be adopted by the industry. - Michael noted we will discuss this today. Opens: - Walter has a proposal to resolve the discussion. Bob Ross suggested one idea is to just start from scratch. Arpad noted he sent an email to the IBIS Interconnect reflector with his thoughts on the direction we could take. Discussion: Touchstone X Discussion: Walter gave an overview of the problems with Touchstone 1.0: impedance per port, port mapping information, size of the S-parameter files, pole-residue format, and s2p port order conventions. He has a proposed scheme to extend the option line. His question is what will make a new Touchstone valuable enough to be adopted in the industry. Walter noted impedance per port was said to be needed for power integrity models, but it is an accuracy issue due to the normalization of the data. Walter noted it is difficult for S-parameter experts to agree on these issues. Walter noted the other issue is the port mapping information. He proposed parseable comments to give the port names. This would allow existing parsers to use the data. Walter commented that the size of the S-parameter files is another issue. What is needed is a factor of 100 reduction in size. He proposed a sparse format, with a utility to convert back to the full format. Walter suggested to make pole-residue a separate specification. He suggested we can reuse the sparse and port mapping information. Walter noted the [Two-Port Data Order] keyword is not strictly followed in Touchstone 1. He suggested, if the keyword is present, the Two-Port mapping is used. Walter noted one Touchstone rule in the current specification is that additional option lines after the first one shall be ignored. He proposed the first option line is the usual Touchstone 1 format. Additional option lines would be keyword lines, such as version, Two-Port Data Order, impedance per port, port name, Sparse info, number of frequencies, network format, and matrix format (e.g. full, lower, upper). Walter suggested to publish the source code of the Touchstone X parser to allow for greater adoption and conversion from internal parsers. Bob noted a lot of these features are already in Touchstone 2, and Touchstone 2 is already supported by EDA tools. If we go a different direction, we would have to grandfather in the Touchstone 2 syntax. Bob commented that there are significant issues with Touchstone 1, and this proposal would be a new branch of Touchstone. Arpad asked if we want the keyword style or a non-keyword style. He asked if the issue with Touchstone 2 is the keyword style or if people do not find the additional required information useful. Arpad was concerned with the down-converter requirement, since as soon as we have per port impedance, we would need to convert the data. Walter confirmed the intent would be to recalculate the data. Arpad noted our existing converter could be extended to do this recalculation. Michael asked if the main objective is to maintain parser compatibility with existing touchstone readers. Arpad asked why we use the pound symbol to denote the keywords. Walter noted that additional option lines after the first one shall be ignored. He suggested it is a better option than using parseable comments. Walter noted the idea with this proposal is that it would work with existing Touchstone parsers. It is only when the information, such as sparse or per port impedance, is needed that the user would be faced with a choice to update the parser or down-convert the file. Walter suggested the required keywords are an issue for adoption of Touchstone 2. Michael gave an AR to everyone to test the proposed option line syntax in their internal parsers [AR]. Arpad moved to adjourn. Randy Wolff seconded. Next Meeting: The next meeting will be March 8, 2023. ================================================================================ Bin List: 1. Draft Touchstone document separating version 1.0 and 2.0 2. Encapsulating Touchstone 1.0 data 3. Touchstone 2.0 draft with TSIRD3 and TSIRD4 4. Pole-residue format 5. Port naming 6. Alternatives to the Touchstone 1.0 option line 7. Enable Cascading of S-parameters Through W-element 8. Touchstone 3.0 draft outline (dependent on several items above)