====================================================================== 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 November 9 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 Minutes: - Michael called for review of the minutes from the November 2 meeting. Bob moved to approve the minutes. Arpad seconded. The minutes were approved without objection. Review of ARs: - Michael to look for a public domain picture. - Michael noted that we will get to this later. - Walter to incorporate his introductory text in the BIRD document. - Michael noted that we will be going over the draft BIRD that includes this text. - Michael to look into the actual name of the Touchstone Reference Port Impedance. - This AR is no longer required since the text in question has been removed. Opens: - Draft 43 of the Interconnect can be reviewed. Michael said he will get it posted soon. - Bob mentioned the passing of Lynne Green (a past Vice-chair of IBIS and co-author of a book covering IBIS). Bob will be sending a message to the IBIS reflectors. Interconnect BIRD draft 43: Michael noted that he will share the BIRD version with the integrated introductory text as draft 43. He noted that are still some required changes from the comments. He fixed a few typos in the examples, but that still needs gone through. Also, the [Manufacturer] examples have been changed from an actual company to a fictitious one. Michael stated that the word "port" has been changed to lower case in the Unused_port_termination subparameter. There was one case where "Package model" was used in place of "Interconnect model" and that has been fixed. Michael noted changes to the Terminal_number section. There was some clarification to the N+1 terminal, as well as some grammatical fixes. Michael commented that there are two open issues including the split models for pin to pad to buffer and the comment on page 22 in regards to Aggressors. Arpad asked if we have resolved the naming of the Aggressor designator, and if we instead wanted to call it "not suitable for victim" which gives a different impression. Michael asked if we can use the comment text in the BIRD text. Bob said that Walter already expanded on this topic in the introduction. Although, he commented that it is still not clear how you choose the victim. The Interconnect model, which includes the coupling information, is not known. Michael suggested to defer the discussion on this until we look into the introductory text as some text may need to be moved around. Bob commented that we may need to clarify a rule that the Aggressor tag applies to the entire path, and we need to think about how to handle conflicting cases. Arpad asked if we have a rule that we cannot have every signal labeled as an Aggressor. He thought at least one should be a victim. Michael noted that the examples need to be gone through in detail, and remaining comment referring to the File_line needs to be resolved. Bob stated that it should be File_IBIS-ISS. He noted that for some of the examples the N+1 terminal should not be reflected in the Number_of_terminals. Michael proceeded through the examples looking at these issues. Randy noted a typo in Example 2. Bob noted that Example 2 uses ground on both sides of the terminals for the IBIS-ISS model. Michael asked if we need another example specifically for Touchstone. Bob noted that you lose conductivity of the ground for Touchstone files, since you have a common reference. Arpad asked if any of the examples use Aggressor. Bob commented that there is one that comes later. Michael stated that Example 8 shows the split pin to pad, pad to buffer case. He commented that we need to scrub the document to remove "Begin" from [Begin Interconnect Model]. Bob commented that one definition of an [Interconnect Model Set] could be a similar group, while another definition is that one grouping can be used for I/O. In the case of I/O, the set must contain the full path. Any pad to buffer only set could cause issues with EDA tools. He stated that everything from pin to buffer should be defined, but we do not have to include all the pins. Walter commented that he thinks it should be up to the model maker what to include in the model. And, the EDA tool should be able to figure out the connections. Bob agreed this is possible for power/ground supplies but not for signals. Michael suggested to move on, and we can resolve this issue when we go through the new introductory text. Michael assigned an AR to Bob to check Examples 10 and 11 and resolve whether they are for IBIS-ISS or Touchstone. Michael commented that there is a rule to be added to the General Syntax Rules and Guidelines that underscore and space can be used interchangeably. Bob noted that this is a guideline to the IBIS specification text. But, we already have this rule for keywords. Michael stated that this rule needs to be clarified. Michael asked about the Rules of Precedence. Bob noted that this is also in the Package section. He expressed concern that these rules are incomplete and there are expected overrides in IBIS that are not explicitly stated. Walter thought that instead of having rules of precedence that we put in a statement saying the Interconnect Models are expected to be more accurate. Michael noted that the new introductory text will be focus next time. He will further look into the existing rules of precedence in the IBIS specification. Arpad moved to adjourn. Bob seconded. The meeting adjourned without objection.