================================================================================ 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 February 12, 2020 Meeting (* means attended at least using audio) ANSYS Curtis Clark Cadence Design Systems Bradley Brim Intel Corp. Michael Mirmak*, Ifiok Umoh Keysight Technologies Radek Biernacki Mentor, A Siemens Business Arpad Muranyi* Micron Technology Justin Butterfield*, Randy Wolff* SiSoft Walter Katz*, Mike LaBonte Teraspeed Labs Bob Ross* Michael Mirmak convened the meeting. No patents were declared. Justin Butterfield took minutes. Review of Minutes: - Michael called for review of the minutes from the February 5, 2020 meeting. Randy Wolff moved to approve the minutes. Bob Ross seconded. The minutes were approved without objection. Review of ARs: - Michael to send out the task list tracking spreadsheet. - Michael reported this was done. Opens: - Randy noted he has a drawing of a Mini-RDIMM with an extended net. - Michael noted that the minutes have not been posted to the website since mid-December and other files have not been posted since August. Arpad Muranyi noted the ATM task group is in the same situation. Michael will discuss this with Steve Parker. EMD Example Diagram: Randy showed the example Mini-RDIMM picture and asked for feedback. Bob asked if the extended net also goes to U1. Randy noted that the extended net goes from the edge to the resistor then to the register. Michael suggested to have an image to show the overall EMD and a couple of zoomed-in images on each of the paths. Randy agreed we could add zoomed-in images of the nets. We could also add an example to show the Terminal lines. Michael noted we want to show, structurally, what the pins connect to and the extended net concept. Randy noted there are components on the back side which are not shown. Michael agreed that this can be difficult to show. Bob commented there are two ways of dealing with extend nets. Randy stated, for the pre-register and post-register, he would not consider these extended nets, since the register acts as a buffer. For the resistor, the nets on each side have different names. Bob commented we do not have a syntax which supports this extended net. Walter Katz noted the syntax can support this by using the extended net name rather than the CAD net for the signal_name. Bob noted each side of the designator would have different signal_names. Walter noted this is true if you have a designator for the resistor, but if you have the resistor embedded in the model, the signal_name can use the CAD nets names. EBD has the same capability, and the EMD syntax can support both of these cases. Walter commented U3 in Randy's example is a buffer and does not create an extended net. Michael noted there was some language in the EMD that passive components can be used, but he could not recall the exact text. Walter noted, for active components, it is up to the EDA tool to handle these, and there can be differences for redrivers and retimers. Michael asked if the user, at a high-level, would know if the components are active or passive. Walter replied the IBIS file will tell you this. Bob noted IBIS was not intended to handle these through paths. Walter stated he is working on the code to create an EMD model. One deficiency in EMD that has come up is how to handle extended nets with passive components vs. active components. Walter stated the issue is when you have components which do not have IBIS models. The EDA tools need ways to handle components which cannot have an IBIS model associated. We can have examples to show how to handle the different cases. Randy noted on page 7 there is some text on this, and he suggested we could add to this. Michael noted there was an AR which he missed for Walter and Randy to create an EMD example based on an RDIMM. Walter stated he has a JEDEC board file, and this is a work in progress that will take some time. It is not necessarily something we want to add to the document, but it is a cross-check if the syntax can work. Walter noted Randy's part is done. IBIS-ISS Parser: Bob requested a quote for the cost and a time line to develop an IBIS-ISS parser. He has not received any feedback. He sent a simplified one page specification for the parser developer. Michael stated we can continue the discussion when we have more information. EMD Editorial Review: Michael noted some of the task list items are simple changes, while others will require more work and discussion. He asked if we want to select an editor and make some of these changes. Walter responded that items #3 and #8 should be easy. Michael suggested to tackle items #7 and #8. Michael asked Bob about item #3 and if we have the text to be replaced in the document. Bob replied this section discusses interactions of terms which have not yet been defined, he suggested to move the text. Bob suggested to discuss the Aggressor_Only rules and have similar pictures to Figures 46 and 47 in the Interconnect section. Michael asked if we need to add references to the Aggressor_Only text in the Interconnect section. Bob replied EMD has similar rules, but they are not the same. There are variations for extended nets that could have different Aggressor_Only tags. Michael suggested to clearly define what needs to be done. Bob stated he needs to work on this. Bob will propose changes to the Aggressor_Only rules [AR]. Randy commented items #6 and #10 are similar efforts and could be combined. Michael agreed. Randy suggested to remove item #6 regarding creating the EMD example of the RDIMM. Michael will make changes in the EMD draft for task list items #6 and #7 [AR]. Michael suggested for Bob to work on the Aggressor_Only rules text. Bob noted this is not on the task list. Michael stated we need a crisp definition of what the task needs to be. Walter commented we need a set of rules for the parser developer. Randy will create zoomed-in versions of the drawings and add some labels [AR]. Next Meeting: The next meeting will be February 19. Walter moved to adjourn. Randy seconded. The meeting adjourned without objection. ================================================================================ Bin List: EMD Comments to be Resolved: (See BIRD202.1 tracking spreadsheet) IBIS-ISS Parser: - IBIS-ISS parser scope document