=============================================================================== IBIS INTERCONNECT TASK GROUP Mailing list: ibis-interconnect@freelists.org ================================================================================ Attendees from May 22, 2024 Meeting (* means attended at least using audio) ANSYS Curtis Clark, Juliano Mologni 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 Weston Beal*, Arpad Muranyi*, Randy Wolff* ST Microelectronics Aurora Sanna Synopsys Ted Mido, Edna Moreno Teraspeed Labs [Bob Ross] University of Illinois Jose Schutt-Aine Zuken USA Lance Wang Michael Mirmak called the meeting to order. No patents were declared. During the minutes review for the May 8 meeting, Arpad Muranyi moved to approve the minutes; Weston Beal seconded. Michael also thanked Randy Wolff for hosting the meeting and taking minutes. The minutes were approved without objection. Michael summarized the completion of his long-standing AR to review the IEEE 370 specification for frequency step size, sampling, and bandwidth information. The specification is split into multiple parts. The content of most interest to the Interconnect Task Group is TG3, Chapter 7, though some material of interest appears earlier. He noted that the port naming convention ordering is odd/even (odd on "left", even on "right"). Further, the specification separates Touchstone data analysis for quality into two phases: - initial or informative evaluation in the frequency domain, using the original data set and reporting causality, passivity, and reciprocity (C, P, R) - application-based normative evaluation in the time domain, using a copy of the data (with the note "[t]he standard does not recommend replacing the original S-parameters") Checking is based on an application-specific data rate, edge rate, frequency resolution, DC value, frequency step, and maximum frequency. The specification uses comparison to an artificial data set with enforced C, P, R. Any evaluation tool will create pulse responses to represent the original data, and will compare the two sets, quantifying C, P, R violations as percentages, in units of voltage and time. All but passivity are defined for both single-ended and differential data. No frequency bandwidth or step size requirement is defined, except in Section 7.4.2, where the step size "should allow for at least eight samples per each 2*pi phase change for the Thru component." In addition, the "portion from DC to the first frequency sample... should not exceed a phase change of pi/2." As noted earlier, Appendix U.2, part 2 defines approximation with rational fitting that closely matches what is already defined in the compression TSIRD from Arpad. The specification also defines a strict port naming format for DUTs and fixtures. Michael quoted page 18: "The port configuration shall be documented in the comment field of the S-parameter Touchstone file as described in 3.5.7. As an important consistency test, it is recommended that the S-parameters always be checked to identify which terms are the insertion loss and verify the assumed port labeling. This is done by examining the magnitude of the loss of the presumed transmission path (e.g., S21 or S31). For a passive interconnect, this is often a small dB value at low frequency." Arpad asked whether Michael had found anything additional in the specification that Touchstone 3.0 would need to support. Michael replied that some document meta data could accompany the Touchstone files in future, containing quality information. But otherwise, he did not think so. Weston stated that quality assessment is a separate issue. He advised treating Touchstone 3.0 similarly to IBIS, where the specification defines the syntax only and separates quality checking. The team reviewed feedback from the IBIS Summit presentation Michael delivered on port-mapping plans. Arpad expressed some disappointment at the lack of feedback. Michael noted that it was at the end of the day for many of the European attendees. Weston suggested that the attendees perhaps did not have strong opinions as the proposed features and methods accomplish the purposes described. In addition, if the participants have not tried linking EMD from Touchstone, then they will not be aware of problems or limitations. Arpad noted that the team had originally started with small list of features for port mapping, which has progressively gotten larger. He was hoping for more feedback on areas (e.g., measurement) where the team had less experience. Michael noted that previous feedback on Touchstone general improvements in 2023 was much stronger, particularly on compression. He asked whether anyone attended this year from oscilloscope or hardware tool vendors. Arpad suggested that the attendees are usually high-level PhDs. Randy stated that there were around 20 attendees in person in the room. He stated that the team should reach out to Donald Telian, Eric Bogatin, and Scott McMorrow to obtain industry feedback across a wider cross-section of users. Michael accepted this AR. [AR] Michael asked whether anything was missing from the priority list shown. Weston noted that the list had "stretched out" far beyond what was originally asked for. Due to conflicts, all attendees but one had to leave after 30 minutes. The meeting adjourned due to lack of attendance. The next meeting will take place on May 29, 2024. ================================================================================ 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 4) Complete pole-residue format BIRD draft (Muranyi) - COMPLETED 5) Complete port naming proposal (Katz) - COMPLETED 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.