====================================================================== 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 August 31 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 Michael Mirmak convened the meeting. No patents were declared. Review of ARs: Review of Minutes: Michael called for review of the minutes from the August 24 meeting. Michael noted a couple of typos in proper names. Mike moved to approve the minutes. Arpad seconded. The minutes were approved without objection. Opens: Michael mentioned that the review of the Interconnect BIRD draft 39 is the main agenda topic. Walter added that the naming of the buffer rail terminals is the only remaining technical issue. Mike agreed and mentioned that he went looking through emails and previous meeting notes, but he was unable to find any discussion or decision on the issue. Walter thought he made a motion to use the same names as [Receiver Thresholds], which are the same as C_comp BIRD, but that motion was tabled and deferred to wait for Radek to make some comments. Michael asked about the rules of precedence and if the changes discussed in last weeks meeting are addressed in draft 39. Mike mentioned that he sent out a bullet list of changes in the email with draft 39. The next items that need to be addressed are the comments, check the examples for relevance and possibly add figures to some of the examples. But, he thought we could move on to the buffer rail terminal naming discussion first. Buffer Rail Terminal Naming: Bob expressed concern about the terminal name change. The C_comp BIRD has different rules from the Interconnect BIRD, so the two should not have the same names for the terminals. It would be good to follow the names with the type such as pin. Mike asked what the proposed names are. Walter answered that they are Pullup_ref, Pulldown_ref, Power_clamp_ref, Gnd_clamp_ref and Ext_ref from the [Receiver Thresholds] convention, and the same names will be used in the C_comp BIRD. These are the five terminals of the [Model] and they should be named the same. [External Model] is different, but we should try to be consistent in the IBIS Spec. He commented that we could open it up for a vote. Bob commented that we follow the terminal name by the pin type for Interconnect modeling, but in the C_comp BIRD we don't follow it with the pin type. A [Model] can serve several different pins so you can't embed the pin name in the syntax. He agreed that the terminal type should be consistent. We had this with using "buf", "pad" and "pin" for terminal type. There is a trade off between inconsistent and clarity. Walter asked if we should vote on it. Bob asked if we have enough people for a vote. Mike mentioned we have about the same attendance as the last few votes. Walter asked if we could schedule a vote for next meeting so that we can get a resolution and move on. Bob seconded that we schedule a vote for next meeting. There were no objections. Walter questioned what string we could add to the names as a qualifier. Michael asked to clarify what we are voting on. Walter said he will send out an email detailing the motion. Interconnect BIRD draft 39 Review: Walter suggested to accept all changes in the document and go through the comments. Comment #1: (In regards to why Parameter is shortened to Param.) Bob noted that the issue is that we overuse the word "Parameter", for instance in IBIS-AMI Parameter is used repeatedly. The word Parameter is used in several other places in IBIS Spec as well. Mike commented that we could add a qualifier to make it more clear. Michael asked if the sentence is clear. Walter requested that if there is a desire to make a change then it should be in the form of a motion. Bob noted that this comment is background information. Michael asked if there was any objection to remove this comment. No objections were made. Comment #2: (More discussion on Parameter vs. Param and relation to AMI) Michael noted that again this is about Param, but that no action needs to be taken. Comment #3: (File names in the Corner syntax) Mike noted that this comment is on the Corner convention. Michael added that this comment is asking if we want to have a Corner for the TS file name. Walter commented that he thought we would handle typ, min, and max corners with different [Interconnect Model Set]s. He noted that Param only supports values, so there is no Corner capability. Arpad commented that we need a mechanism to point to different subckts in the file. Walter noted that on page 14 this is addressed in the File_IBIS_ISS syntax where the subckt name is passed in as the second argument. Mike asked if should we delete the entire paragraph related to the File_TS and Corner syntax. Bob commented that we could replace it with some better text to resolve this. Comment #4: (Z0 and TD discussion) Michael noted that this is a very early comment that is outdated, since the impedance variation is deleted. He stated it is safe to delete this comment. Comment #5: (Asking if keywords should have brackets) Mike noted this was his comment and it is still an open issue as to whether keywords should have brackets in the document text. Arpad commented that we usually do use the brackets. Bob noted that in this case we are introducing a keyword that hasn't been defined yet. So we can add a forward reference after the keyword. Mike will take an AR to add brackets to all keywords. Arpad and Bob agreed. Bob reiterated that we should add a forward reference since [Interconnect Model Set] has not yet been defined in the document. Michael suggested to use "defined elsewhere" as the forward reference text. Comment #6: (Asking what the file name should be for [Interconnect Model Set Selector] when it is defined inside the .ibs file) Bob asked if we are we using .ims or .ict for the [Interconnect Model Set] extension. Mike commented that we seem to use both in the document. Bob thought we should be consistent. Michael asked if this was a new change. Mike said that it was new. Walter moved to change all instances of .ict to .ims in the document. Arpad seconded. There were no objections. Walter asked if we should we allow the ability to use an [Interconnect Model Set] from another .ibs file. Bob said he would be against this. Arpad asked why would you want to do that. Michael commented that you could a have a separate .ibs file with the [Interconnect Model Set] information. Arpad noted that we have never done this before. Bob thought it should not be allowed. Walter said he was okay with preventing it. Mike noted the comment (#6) is about the syntax of the *.ibs after the [Interconnect Model Set] name to denote the file. Walter commented that we have three choices: *.ibs, nothing or NA. Bob mentioned he was okay with *.ibs and thinks that this is clear. Arpad proposed to use "In Line" instead of *.ibs. Walter moved to change the syntax such that if the [Interconnect Model Set] is defined in the same file then the second field would be blank. Michael asked if there was there a second. Since there was no second, the motion did not carry. Arpad asked how we do this with the .pkg file name. Bob answered that the default is to search, but he recommended to not use that convention. Arpad agreed. Arpad made a motion to change the text to a more descriptive sting such as "In Line". Walter seconded. Bob expressed concern over "In Line", as it would be a reserved word or argument. Mike commented that the key trait of the name is that it would not conflict with an actual file name. Walter suggested to use the single character '.'. Michael said we will have to continue the discussion next time. Michael took an AR to phrase a vote on the buffer rail terminal naming to email reflector. Bob noted that he has more comments on the draft for the next meeting. Walter moved to adjourn. Arpad seconded. The meeting adjourned without objection.