============================================================================== IBIS INTERCONNECT TASK GROUP Mailing list: ibis-interconnect@freelists.org ============================================================================== Attendees from September 24, 2025 Meeting (* means attended at least using audio) ANSYS Curtis Clark, Wei-hsing Huang, Juliano Mologni Arista Networks Jim Antonellis Broadcom James Church Chipletz Stephen Newberry Intel Corp. Michael Mirmak*, Xiaoning Ye Keysight Technologies Ming Yan Marvell Steve Parker MathWorks Walter Katz* Micron Technology Justin Butterfield Siemens EDA Weston Beal*, Arpad Muranyi*, Randy Wolff* Simberian Yuriy Shlepnev ST Microelectronics Aurora Sanna Synopsys Ted Mido, Edna Moreno University of Illinois Jose Schutt-Aine ============================================================================== Michael Mirmak convened the meeting. No patents were declared. Michael reviewed the minutes of September 17. Randy Wolff moved to approve the minutes; Arpad Muranyi seconded. The minutes were approved without objection. Michael noted that all ARs have been closed. During review of the Port Mapping proposal, Arpad detailed several concerns with Draft 19: 1) He prefers the spelling "subparameter" to "sub-parameter". Randy agreed. 2) He suggested that an example of Swath is needed (it appears three times in the document). Michael added that this should likely be a connector with a schema reference, but the example will have limited value as no schema definition has yet been established and it won't be part of this version of Touchstone. Arpad agreed, and suggested adding a comment about schema files. Weston Beal noted that swathing will be most common for connectors, with edge, end, and center sections. Arpad suggested that flat ribbon cables would also be likely candidates for swathing. Weston agrees. Arpad added that this would be similar to press-fit IDC (Interconnect Displacement Connector). Weston noted that establishing the swath details may be difficult. 3) Arpad raised concerns about units ordering; can we group them into metric, then non-metric? Michael asked whether we need a default. Arpad replied that it's an optional parameter, so no. Michael added that the parser will flag when Physical is used without units. Arpad suggested that metric units be placed first, with imperials next. This was done in-line in Draft 20 during the meeting. 4) The Port section is written somewhat vaguely. Arpad feels that a better approach would be starting with "1" and covering all numbers (Michael suggested these should be integers) through N inclusive, where N is the Number of Ports. This was done in-line in Draft 20 during the meeting. 5) Arpad asked whether the Ports definitions are required to be in a single line, or whether line wraps supported. Michael asked whether this implies a need for a line continuation character. Arpad replied no, as we have the Port keyword to indicate this. Michael added that other subparameters also act as terminators. Weston noted that we could require all Port contents in one line; if a person is looking at it, their editor can handle it. Arpad asked what should be done for EMD and IBIS Interconnect Weston replied that there isn't as much data there. 6) Arpad suggested making the transistor labels in the diagram larger (the current diagram looks as if it is using a font size of 9). 7) In the transistor example, Arpad suggested removing port 3. Walter Katz raised questions about the labels used in the transistor example. Randy suggested labeling a port as the reference. Michael replied that there is no way to label the Emitter *node* without it being in a *port*, as each port is a pair of nodes. Michael asked whether we can "smuggle" a terminal into the drawing and definition. Arpad suggested that the Logical label can be used in this case. Michael suggested that, predicting the future, we will forget this context if we add Logical as a label here. Randy asked whether the specification should disallow Logical without Physical being present. Arpad replied asking what therefore is the meaning of Physical. The model needs to tell the EDA tool about connectivity. Michael replied that connectivity is the problem; this Logical-only example can't be connected based on the names. Randy disagreed with this. Walter suggested this was an s3p file, though it does not serve as a great example for a schematic symbol. Michael noted that an early Hewlett-Packard application note did not use ground as a symbol anywhere. In addition, we have created two different uses of schematics, including for connectivity as well as for symbol creation. Randy replied that measurement data is different than symbol definition. We should show the symbol in a schematic with labels such as left_side port 1, top_side port 2, etc. Arpad suggested showing how the data was measured in two ports, with an AC termination to a bias source. Arpad suggested that providing two cases using the same transistor structure; the second case may be shown twice, once as a real circuit (3-terminal SPICE) and once as an S2P structure. For Michael's second drawing, showing board interconnect, Arpad suggested making the caution statement larger and adjusting the lines to be identical. He also suggested changing the Interconnect example to an EMD Model example. Arpad noted that A_gnd is missing from several cases; we need to add examples with A_gnd, plus complete a comprehensive example review. We also probably need to remove Example 15's A_gnd node references. Arpad moved to adjourn the meeting; Randy seconded. The meeting adjourned without objection. ==============================================================================