====================================================================== IBIS INTERCONNECT TASK GROUP http://www.ibis.org/interconnect_wip/ Mailing list: ibis-interconnect@freelists.org Archives at http://www.freelists.org/archive/ibis-interconn/ ====================================================================== Attendees from July 20 Meeting (* means attended at least using audio) ANSYS Curtis Clark Cadence Design Systems Bradley Brim Cisco David Siadat Intel Corp. Michael Mirmak Keysight Technologies Radek Biernacki, Ming Yan Mentor Graphics Arpad Muranyi* Micron Technology Justin Butterfield, Randy Wolff* SAE ITC Maureen Lemankiewicz, Logen Johnson Signal Integrity Software Walter Katz*, Mike LaBonte* Teraspeed Labs Bob Ross* University of Aveiro in Portugal Wael Dghais Mike LaBonte convened the meeting. No patents were declared. Review of ARs: AR: Bob Ross and Walter Katz edit Terminal_type association table. This was done, but Bob Ross said a final decision was still needed. Mike LaBonte noted that several proposals had been submitted by email. Bob said he was withdrawing his crosshatch format proposal. Review of Minutes: The attendees reviewed minutes from the July 13 meeting. Bob Ross moved to approve the minutes. Walter Katz seconded. The minutes were approved with no objections. Terminal Names [Receiver Threshold] format discussion: Bob Ross noted that Radek Biernacki had an interest in this topic, but was not present. Walter Katz moved to table the agenda item just for today, Bob Ross seconded. The motion passed with no objection. Updated Terminal_type association table: Mike LaBonte showed "More Interconnect BIRD Terminal_type Association Table Formats" and and gave an overview. Mike favored proposal D, which had separate I/O pin_name and Rail pin_name columns. Bob Ross said the "X" in the rail pin_name column really needs to be an I/O pin name, and that he would reject proposal D because it is missing information, having only "X" characters in all cells. Walter Katz noted that he had sent a proposal in an email July 13, and he showed his proposal. Walter said the characters used in the table cells could change, but but overall he felt the format was simple and clear. Bob said "W" might be confused with W-line. Mike asked why the letter "W" had been chosen. Walter said it was simply that "W" was the letter before "X" in the alphabet. Mike suggested the table might be more clear if short words with more than one letter were used in the cells. Bob said we we would want to avoid using words that could be confused with other similar terms in the specification. Mike LaBonte showed IBIS 6.1 page 230. Mike noted that we had been using only "X" values in the cells of "allowable usage" tables. He asked if this was a different kind of table, such that it warranted more detail. Bob Ross said there was a need for more information. Walter Katz shared the table from his email again. Mike LaBonte suggested replacing "X" and "W" with something more descriptive. Walter changed all "X" to "I/O" and all "W" to "Rail". Bob Ross suggested that "S" could be "Rail". Walter said "Rail" was already used to denote pins, so it should not be used for signals. Mike suggested "S" could be "Sig" and "B" could be "Buf". Walter made that change. Walter Katz moved to adopt the Terminal_type association table as edited during the meeting. Bob Ross seconded, commenting that we might revisit "Die" at some future point. The motion passed without objection. Walter said he had already sent the table just adopted to the list, no need for an AR. Interconnect Model Sets requirements: Mike LaBonte showed a table depicting items from the existing [Define Package Model] scheme and their equivalence to items in the new [Interconnect Model Set] scheme. This had been sent in a email the prior week. Bob Ross agreed that the table matched his perception of the equivalences. Walter Katz said [Interconnect Model Selector] should be [Interconnect Model Set Selector]. Mike made that change. Walter said that with hindsight the old keyword might have been [Package Model Selector] instead of the [Package Model] and [Alternate Package Model] set. He said an [Interconnect Model] was equivalent to a section of a [Define Package Model]. Walter noted that in the old syntax a [Pin Numbers] section could not be coupled. Bob Ross said [Number of Sections] denotes the number of stubs, but this still did not signify any coupling. Walter Katz showed IBIS 6.1 page 156, where [Path Description] is described for EBD files. He said [Number of Sections] there had nothing to do with [Number of Paths]. He showed page 141. He said it was really [Pin Numbers] that established the interconnect in a [Define Package Model], and that that was analogous to [Interconnect Model]. Walter said he had chosen not to propose a scheme like [Alternate Package Model. Bob Ross said [Pin Numbers] was used for either a single power pin or a single line I/O pin. He said the question was whether there should be a rule that every [Interconnect Model Set] should contain at least one [Interconnect Model] which contains an I/O pin. Mike suggested this should be the first topic for next week. Walter Katz moved to adjourn. Bob Ross suggested that we should decide in the next meeting if the models in an [Interconnect Model Set] fully describe the package, then Bob second the motion to adjourn. The meeting adjourned without objection.