====================================================================== 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 24 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: Review of Minutes: Mike called for review of the minutes from the July 27 and the August 17 meetings. For July 27, Arpad move to approve. Walter seconded. The minutes were approved without objection. For August 17, Walter move to approve. Arpad seconded. The minutes were approved without objection. Opens: Mike mentioned the interconnect BIRD draft 38 was sent out yesterday (August 23). Interconnect BIRD draft 38 difference document: Mike had started with draft 37 of the BIRD and went through the bullet list of changes he sent out to the group. He made BIRD template changes for consistency. The instructional text was made blue. He changed [Interconnect Model Selector] to [Interconnect Set Model Selector]. He moved [Interconnect Model Set] description from near [Interconnect Model Set Selector] to near [Interconnect Model] in the document. In the [Interconnect Set Model Selector], the Other Notes section, giving the rule requiring a full path, has been moved. Mike asked if we had agreed to remove the [Bus Label] keyword. Walter disagreed and mentioned he thought we should keep the [Bus Label] keyword structure. Mike said he will fix it in the next draft. Walter commented that the reason for needing [Bus Label]s is to connect the component together. If you want to split up sections of Interconnect, say VDD or VSS with same name, [Bus Label] is required. If VDD is split on the silicon, [Bus Label] is necessary to connect the rails, but the component pins might not have the same split. The [Bus Label] syntax allows for the grouping. Mike asked why the [Bus Label] keyword is necessary when we have [Pin Mapping]. Walter gave an example of when you only have one VDD pin to connect to multiple split sections of VDD on the die. [Bus Label] allows you to connect the VDD pins with how the pads are organized on the silicon. Mike said that makes it clear and agreed we should keep the keyword, and he will update it in draft 39 of the BIRD. Mike commented that we need to make changes to the Package Modeling section in the IBIS Spec to expand it beyond the [Define Package Model] keyword. We need to add some introductory text to clarify the scope of Chapter 7 versus the new Chapter 13 for Interconnect Modeling. Mike added that for the new Chapter 13, he made changes to the introductory text and added entries to the Interconnect Modeling Keywords and Subparameters table. The [Interconnect Model Set] definition text is now merged with Walter's changes. He aslo removed "begin" from the [Interconnect Model] keyword. There was also a comment about aggressors that he moved from document text to the document review comments. He also edited all of the examples to remove "begin" from keywords and to enclose the [Interconnect Model]s in the [Interconnect Model Set]s. Arpad asked about the nested hierarchy with keywords encapsulating other keywords. And, how to handle this if we have encapsulated sections versus encapsulated subckts in Spice. Walter commented that in Spice it is a scoping issue with subckts. The only thing you can have inside a [Interconnect Model Set] is [Interconnect Model]s. The "end" keywords are technically unnecessary, but make parsing easier. Arpad asked if we should have keywords contain other keywords. Mike gave an example of [Rising Waveform]/[Falling Waveform] keywords where you can have multiples inside a single [Model] as a precedence for this type of syntax. Arpad agreed that maybe it is not that big of a concern. Rules of Precedence: Aprad asked why Chapter 13 is defined before Chapter 12 in the BIRD document. Mike responded that Chapter 13 is the main topic of the document Walter noticed, in the first paragraph of the 12.1 Packages section, we should change the wording from "that data shall be used..." to "that data shall be considered be more accurate" or some along those lines. This should be done in order to not force the EDA tool to use one type of model over the other. Mike agreed. Walter thought maybe the Rules of Precedence chapter should be moved to the beginning of the IBIS Spec. Arpad asked if it should be in Chapter 7 instead since it pertains to packages. Mike thought that we should look into it and added a comment to draft 38 to capture these observations. Walter commented that on page 20 of the IBIS Spec 6.1 in the [Package] keyword Other Notes there are some rules of precedence already. Mike noted that the [Package] keyword needs to be updated, and we need to look at the text for all the package keywords. Walter thought that rather than update all of the keywords we need new rules of precedence text. Mike added the current rules of precedence should also be in the interconnect modeling section. Walter agreed. Arpad added that we also need to clarify [External Circuit], since it is mentioned in the current Rules of Precedence text in the BIRD. Mike said he will go in and make small changes to draft 39, and asked if there was enough consensus to make changes to Rules of Precedence in draft 39. Walter said he thought there was, and he added that these are our only tasks left minus editorial work and possibly removing some examples. Mike suggested that we also have a linear review going through the document from beginning to end. Walter agreed. Mike questioned about adding more rules of precedence for other IBIS structures such as [Model Spec]. And he thought we could swap the chapter numbering and have Rules of Precedence at the end of the IBIS Spec. Arpad commented that if this is to be a general rule, it should be in the beginning of IBIS, possibly in chapter 3. Mike agreed. And he said he will take a shot at making these changes and add rules of precedence in Chapter 3 as a section 3.2 after the Keyword Hierarchy section 3.1. Arpad mentioned he would like to get a somewhat final draft that we could read through. He would like to send it out to additional people to get their comments and also have a more final review. Mike also added that we need to check on some of the comments in the document and resolve them. Walter added that we need to clean some of these comments, so we can focus on the ones that need to be resolved and looked into. Mike stated that he will create a draft 39 with these changes. Arpad moved to adjourn. Walter seconded. The meeting adjourned without objection. Summary of ARs: