IBIS Macromodel Task Group Meeting date: 12 May 2020 Members (asterisk for those attending): Achronix Semiconductor * Hansel Dsilva ANSYS: Curtis Clark * Wei-hsing Huang Cadence Design Systems: * Ambrish Varma Ken Willis * Jared James Intel: * Michael Mirmak Keysight Technologies: * Fangyi Rao * Radek Biernacki Ming Yan * Todd Bermensolo Stephen Slater Marvell Steve Parker Mentor, A Siemens Business: * Arpad Muranyi Micron Technology: * Randy Wolff * Justin Butterfield SiSoft (Mathworks): * Walter Katz Mike LaBonte Teraspeed Labs: * Bob Ross Zuken USA: * Lance Wang The meeting was led by Arpad Muranyi. Justin Butterfield took the minutes. -------------------------------------------------------------------------------- Opens: - None. ------------- Review of ARs: - Arpad to add BIRD198.1 discussion to the agenda. - Done - Hansel to email his updated BIRD draft and presentation to the ATM list. - Done. - Arpad to ask Steve to upload Hansel's documents to the ATM work archives. - Done. Arpad confirmed these are available on the ATM website. - Walter to work on a BIRD201.1 draft to address Radek's feedback. - Done. -------------------------- Call for patent disclosure: - None. ------------------------- Review of Meeting Minutes: Arpad asked for any comments or corrections to the minutes of the May 5 meeting. Michael moved to approve the minutes. Walter seconded the motion. There were no objections. ------------- New Discussion: BIRD198 Keyword additions for On-Die PDN Modeling: Randy shared the latest BIRD198 draft, which introduces a simplified PDN model consisting of a two resistor and one capacitor model. Randy received the draft about a month ago. He has made editorial changes and comments. He requested volunteers to read through the draft and provide additional comments to the authors for issuing a BIRD198.2. Most of the changes from the original BIRD198 to BIRD198.1 were detailed at the DesignCon IBIS Summit. Randy's intent was to not change any technical content. Randy noted some of the keywords were grouped together in the descriptions, and these have been separated to be more consistent with the IBIS specification. He rearranged the text in some sections and clarified the language. Arpad asked about the capitalization of "Bus_label". Randy noted there is a difference between the parameter "Bus_label" and the usual concept of "bus_label", but this could use further review. Randy changed the word "nodes" to "terminals". He tried to follow some of the language used in the Interconnect Model section. One of the figures also needs to be changed from "nodes" to "terminals". He asked if there are any other volunteers to take a look at it. Randy will send the BIRD198 draft to the ATM list. Discussion on “Gap in IBIS for sampling with statistical mode AMI models”: Hansel shared the latest BIRD draft. Fangyi asked about the Usage Rules, where the Rx_Decision_Time would take precedence over Rx_Clock_Recovery_Mean. And, will Rx_Clock_Recovery_Mean be ignored when Rx_Decision_Time is used. Hansel replied the Rx_Clock_Recovery_Mean comes into play when the model does not return clock_times. To keep consistent with the IBIS specification, the Rx_Decision_Time will take precedence. Walter noted Rx_Decision_Time will be ignored for time domain, but Rx_Clock_Recovery_Mean will be ignored for statistical. Arpad commented this is noted in the BIRD. Bob suggested to put Hansel as the requestor. Hansel stated he needs to ask permission from the co-authors. Arpad asked if we are ready to submit this to the IBIS Open Forum. Walter stated we can go ahead and submit unless there are any additional comments. Walter moved to submit the Rx_Decision_Time BIRD draft to the IBIS Open Forum. Bob seconded. There were no objections. Hansel to send the Rx_Decision_Time BIRD draft to Randy for submission to the IBIS Open Forum. BIRD201 Back-channel Statistical Optimization: Walter made changes per Radek's suggestions, and he requested Radek to review these. The changes relate to the transition between the training states. Bob had also suggested editorial changes, and Walter requested Bob to make the changes. Arpad asked if these are still required. Bob replied he has cleaned up the table, where we have a new template for the table. Bob asked if the new DLL function signatures would be a new subsection in 13.9. Arpad asked if Bob can make his changes and send this out for Radek to review. Radek noted he still has comments on the draft. Bob state he has questions relating to the titles and the subsections of the text to be added, and he would like clarity on this. Walter noted most of the changes were in the section about the transitions. He also replaced the table YY1, to match Bob's example. Bob noted the BCI_Training_Mode is missing from this table now. Walter added this line, and listed the default as "GetWave". Bob suggested to change the "Yes if ..." to a footnote text to match the other parameters. He also suggested to add a note that this parameter is for IBIS 7.1. Radek suggested to change "Impulse" to "Statistical" in note 5. Bob suggested for BCI_Protocol to have no default. Bob asked about the default for BCI_Training_Mode. Walter noted the default is "GetWave", which is a valid entry for the parameter. Arpad noted some editorial fixes to the table. Bob commented the fonts in the table are not consistent. On page 2, Bob asked about the statement regarding the BCI_Training_Mode being illegal before AMI_Version 7.1. Walter replied the table and the text are saying the same thing, but in a different way. Radek suggested to make the wording consistent with the BCI_Protocol parameter. Bob suggested to make it say illegal before 7.1. Arpad suggested to leave this as is and make it consistent with the rest of the IBIS specification. Bob suggested to change "DLL" to "executable model" in the BIRD text. Walter made this change. Bob asked about the titles and how the subsections should be labeled. Walter stated he is okay with how we want to organize it. Bob commented he can make the changes to this. Arpad asked if this is one section or not. Walter noted the different pieces might go different places, such as the flow descriptions and parameter definitions being in different sections. This is an editorial issue. Radek commented this text goes into the section 10.9, and it may go in various places in that section. - Michael: Motion to adjourn. - Radek: Second. - Arpad: Thank you all for joining. AR: Randy to send the BIRD198 draft to the ATM list. AR: Hansel to send the Statistical Clocking BIRD draft to Randy for submission to the IBIS Open Forum. ------------- Next meeting: 19 May 2020 12:00pm PT ------------- IBIS Interconnect SPICE Wish List: 1) Simulator directives