IBIS Macromodel Task Group Meeting date: 22 February 2022 Members (asterisk for those attending): Achronix Semiconductor: Hansel Dsilva Amazon: John Yan ANSYS: Curtis Clark * Wei-hsing Huang Cadence Design Systems: * Ambrish Varma Jared James Google: Zhiping Yang Intel: * Michael Mirmak Kinger Cai Alaeddin Aydiner Keysight Technologies: * Fangyi Rao * Majid Ahadi Dolatsara Ming Yan Radek Biernacki Rui Yang Luminous Computing David Banas Marvell Steve Parker Mathworks (SiSoft): Walter Katz Mike LaBonte Micron Technology: * Randy Wolff * Justin Butterfield Missouri S&T Chulsoon Hwang Siemens EDA (Mentor): * Arpad Muranyi Teraspeed Labs: * Bob Ross Zuken USA: Lance Wang The meeting was led by Arpad Muranyi. Randy Wolff took the minutes. -------------------------------------------------------------------------------- Opens: - Michael had a quick question on the root name BIRD needing feedback before he sends out draft 3 to the email reflector. ------------- Review of ARs: - Walter to send out BIRD213.1 draft 13 incorporating Feb. 15 changes. - Done. -------------------------- Call for patent disclosure: - None. ------------------------- Review of Meeting Minutes: Arpad asked for any comments or corrections to the minutes of the February 15th meeting. Michael moved to approve the minutes. Randy seconded the motion. There were no objections. ------------- New Discussion: BIRD213.1 draft 13, PAMn: Arpad shared draft 13. Arpad reviewed the BIRD last night. Randy had found some typos to fix as well. Arpad shared an email chain including comments from Fangyi, Arpad, and Randy. Arpad edited a Draft 14. Fangyi agreed with Arpad's comment about combining discussion of NRZ and PAM4. Arpad copied his new text into Draft 14. Fangyi agreed about Arpad's comments on nominal eye. Fangyi thought some statements in Other Notes should be in Usage Rules in the PAM_Offsets section. Fangyi suggested clearing up the language on nominal eye and replace the entire Usage Rules section with suggested text from his email. There are separate descriptions for time domain and statistical simulations. In the statistical simulation description, Arpad asked if we needed to have an equation to define the nominal eye. Fangyi noted you want the middle eye because that one is usually the best one to reference. Ambrish asked why we need the offset for time domain simulation. Fangyi said each eye has its own slicer, and with only one clock times returned, you need to know the sampling time offsets of the other slicers that are not the nominal eye. Arpad asked if you should have separate jitter for each eye slicer. Fangyi did not think the slicer would add much jitter to the sampling time. Fangyi noted that for threshold voltage levels, you move sampling location vertically, while for offsets you are moving the sampling time horizontally. Walter joined and asked why we needed to define the nominal eye for statistical simulations. Fangyi noted there is no clock times in statistical simulations. Fangyi said we need a definition of t0 if Rx_Decision_time is not present. The first and third paragraphs of the old Usage Rules section were deleted. Randy thought a sentence about positive and negative values that was deleted seemed important to clarify in the new text. Fangyi noted a description did exist for the time domain section but needed to be added for the statistical simulations paragraph. This was corrected. Arpad noted in his new text there was still a sentence with "the electrical interface to either the driver or the receiver is differential". This was edited to "the electrical interfaces to both the driver and the receiver are differential". Two instances of "Parameter" were changed to lower case. AR: Arpad to send out the BIRD213.1 draft 14. Root name BIRD discussion: Michael said for AMI_parameters_out, the EDA tool can compare the root name returned by the DLL to the root name in the AMI file. Arpad confirmed his tool does this. Michael said for AMI_parameters_in, the match could only be done by the model. Michael noted a clarification in the BIRD added that "The EDA tool must report any root name mismatch detected, but may choose to continue to stop simulation at this point." Arpad asked if you have only "in" parameters, does the model have to return a blank string with just the root name. In that case the EDA tool could check the root name. Michael highlighted an existing sentence on page 225 describing the AMI_Init function will return nothing but the root name if there are no out parameters to output. Michael wanted to confirm that exercising the EDA tool is the only way to check for the root name in the output string. Walter and Randy agreed with this. - Michael: Motion to adjourn. - Ambrish: Second. - Arpad: Thank you all for joining. AR: Arpad to send out the BIRD213.1 draft 14. ------------- Next meeting: 1 March 2022 12:00pm PT ------------- IBIS Interconnect SPICE Wish List: 1) Simulator directives