=============================================================================== IBIS INTERCONNECT TASK GROUP Mailing list: ibis-interconnect@freelists.org ================================================================================ Attendees from September 11, 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. During review of the September 4 meeting minutes, Arpad Muranyi requested the following changes: 1) Please add the highlighted word in: “when file permissions *are* not available”. 2) Change the following statement from: "+ Physical and - Physical, for polarities" to: "+ Physical and - Physical, for being able to describe the connectivity of the two terminals of a port independently" Walter Katz would prefer the nodes be described as complementary, noting that for Touchstone, the polarity is meaningless, but that for IBIS, it is meaningful. Arpad replied that he does not have a preference for what specific new language is used. Arpad moved to approve the minutes with these changes; Lance Wang seconded. The minutes were approved without objection. Michael noted that GitHub is being used to file requests for IBIS web page updates, and that he had filed requests to post the latest port mapping drafts. Randy pointed it out as an "interim solution" and a tracker for Steve Parker to ensure web updates are completed. Lance added that he is driving to set up a web interface for the Task Group leaders to upload documents directly. Arpad shared examples of his port mapping proposal. He referred to a March 27 presentation and noted that he is open to suggestions and corrections. Arpad provided two examples for on-die interconnects and packages; the other seven examples relate to EMD. Several items need updating: - need units for xyz coordinates in the port mapping definitions - parameter names are too verbose (need Ref for Reference and similar abbreviations for Physical, Logical) - need complementarity syntax ("Net") as ports may not share the same name within a pair - need wildcards - need syntax for all possible IBIS and EMD connectivity options Walter responded to the recommendations: 1) On units, some databases combine units; he advocated using a single unit definition for the entire data set 2) Walter agreed with Arpad on the verbosity of the parameter names 3) On the negative terminal of a net, Walter suggested examining example 4, ports 163 etc. where Net VDD has a "gnd" reference. Reference is a physical thing (xyz location, refdes pin number). Arpad: noted that the wildcard * is an EMD syntax for pin name. Walter observed that GND is not a bus label, but a signal name. Arpad agreed, noting that syntax refinement needed here to distinguish pin, bus label, and signal name. Walter added that bus_label identifies bus labels, and that the string "Group" followed by group name. Physical is the default. Arpad clarified his concern: GND is connected to something in the CAD database and that something will NOT be VDD. Terminals are connected to two different net names. GND could be a signal name or bus_label name. Randy noted that the syntax *.GND shorts all refdes grounds together. Michael asked about references for differential ports. Arpad noted that we already declare positive vs. negative terminals Randy added that GND is meaningless without an explicit group definition. Arpad replied that we are trying to make a connection definition, not a measurement. Interconnect definitions by_pinname, by_padname, etc. will be needed in the port mapping draft. Walter and Randy replied that we have defined physical names, but they may not be complete for IBIS Interconnect, EMD, etc. Arpad took the AR to update his examples with corrections. [AR] Arpad noted that EMD cannot define a separate reference on each interface. Randy added that the same issue is true of IBIS Interconnect. Michael asked whether EMD fixes are needed and whether Arpad was addressing this. Arpad replied that he was, but that we could also define a port mapping syntax now and later make the update to EMD. Randy noted that per-pin referencing would be needed, to allow connections of IBIS-ISS to Touchstone. Walter noted that bus labels are only involved when bus_label is used. Arpad replied that EMD has two pin lists which may implicitly define bus labels. Walter stated that GND is a legal bus label; a bus label can be a signal name, but they are always bus labels as well. Randy noted that the issue is per-port referencing; you cannot create a terminal list in EMD that supports this. The fact that you have four different references is illegal. Michael took the AR to add "A_gnd" as a reserved and legal term for the IBIS Interconnect and EMD structures in the port mapping draft. [AR] Arpad asked what the equivalent of CONNECT would be. Randy replied that it is the Net parameter. Michael suggested that the examples need diagrams, to which Randy replied that the EMD portion of IBIS already has extended net syntax and pictures to illustrate the connectivity. Walter, after some research, confirmed that a Touchstone file in EMD only has one reference node. Arpad took the AR to update the "Physical" example presentation and syntax. He will provide some changes in an "AM2" draft for Michael to create draft 10 of the port mapping proposal. [AR] Arpad moved to adjourn the meeting. Randy seconded. The meeting adjourned without objection. The next meeting will be held September 18, 2024. ================================================================================ Bin List: 1) Complete port naming proposal (Katz et al) 2) Complete/revise Touchstone 3.0 draft outline (Mirmak) 3) Complete ISS-IRD 1 Draft - enable cascading of S-parameters through W-element (Mirmak) - TABLED Tabled ARs: - Arpad to give an example of the physical connectivity needed for EMD automation.