IBIS Editorial Task Group Minutes October 5, 2022 Attendees: Intel Corp. Michael Mirmak* Micron Technology Randy Wolff* Siemens EDA Arpad Muranyi*, Weston Beal Teraspeed Labs Bob Ross* Minutes by Randy Wolff -------------------------------------------------------------------------------- Call for Patents: none Review of Minutes: Minutes from September 28, 2022 were reviewed. Arpad moved to approve. Bob seconded the motion. Minutes were approved. Review of Actions Required: Michael reviewed ARs from the September 28, 2022 minutes. Michael to research if a collection of sections in individual documents could later be collated into a larger final document [AR] - Michael will discuss more in the main agenda. (to remain open) Arpad to prepare a version of the BIRD draft that modifies the clock_times text already edited in the 7.2 draft document [AR] - done Agenda: Review Linked Document Structure: Michael got the last two chapters and the main document structure stripped down with the changes accepted. He has not finished the full main document. Some changes required letting the Word document sit overnight. He was working to identify chapters where images need compression. Chapters 6, 8, and 11 need the most work to reduce image sizes, with chapter 6 on buffer modeling being the priority. Michael noted change tracking includes comments and changes to 7.2 content. The process deletes chapters with change tracking on, then accepting the change of the deleted text is what takes a long time with some chapters. Michael has not gotten to the point of finding out if figure numbering works in the re-combined document. Randy asked if the size of images in a chapter could be investigated to find figures that need compression. Michael found most images in Chapter 6 are Visio images. Michael will try to convert the Visio images to a a compressed image format such as JPEG [AR]. Review Clock Times Clarifications Updates: Arpad started a BIRD draft 0.3 document. He edited a sentence to remove language about "waveforms being symmetric around the horizontal axis". Walter commented on this text in the ATM task group meeting. Randy and Michael agreed to the change. The group discussed the comment from Ambrish about removal of the text "If there are no valid clock ticks for the duration of an AMI_GetWave call, a single entry of -1 will be returned in the clock_times vector." It was decided the new text was a better rewording of the original text, and the removed text was redundant. The group reviewed the comment from Walter that the added text on waveform continuity was obvious and not needed. Waveforms are defined on page 228. Arpad added this text for consistency with already included statements about clock times monotonicity requirements. Randy commented that he could see Walter's point that the waveforms could not get screwed up based on the existing requirements of block size and sampling times. The existing text about clock times was necessary to point out the issues that can happen with not having enough data precision. The text was removed. Arpad reviewed a comment about the term "next" for continuity of clock samples between AMI_GetWave calls. Michael and Randy thought this was correct usage of that word. Arpad reviewed a comment about whether text should be added describing sampling issues with DQS and DQ. Randy thought we needed to gain more experience using these clocked models before adding any statements about potential issues. We may need to add such statements in a future version of IBIS. Arpad asked whether the text "If an Rx AMI_GetWave only returns -1 during all AMI_GetWave calls then the model does not generate clock times." was clear enough to indicate this applies to models with clock_times as an output. He will consider further if this needs to be changed. Arpad will save the document as draft 3 and send it to both the Editorial and ATM task group emails for feedback and review [AR]. Next Meeting Scheduling: October 5, 2022 Arpad moved to adjourn. Bob seconded. Meeting adjourned at 9:02 a.m. PT.