IBIS Macromodel Task Group Meeting date: 22 March 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. Curtis Clark took the minutes. -------------------------------------------------------------------------------- Opens: - None. ------------- Review of ARs: - AR: Arpad to send draft 18 of BIRD213.1 to the ATM list. - Done. -------------------------- Call for patent disclosure: - None. ------------------------- Review of Meeting Minutes: Arpad asked for any comments or corrections to the minutes of the March 15th meeting. Randy moved to approve the minutes. Ambrish seconded the motion. There were no objections. ------------- New Discussion: BIRD213.1 draft 19, PAMn: Arpad shared draft 19. He had looked through IBIS 7.1 to find mentions of the existing PAM4 parameters, possible contradictions, etc. Based on that investigation, he had emailed ATM a list of suggestions and created a draft 19 including them. On page 273, in the Description of Rx_Receiver_Sensitivity, the existing PAM4 Threshold parameters are mentioned. Arpad had added the new PAM_Thresholds parameter to the sentence. The group agreed with the change. Ambrish asked whether we could consider making the existing PAM4 parameters illegal, since the new PAMn support parameters are being added. Bob, Walter and others said that making the PAM4 parameters illegal is not possible because of the longstanding IBIS policy of maintaining forward compatibility (you can take a file with an older version and increase the version value and still have a legal model). Ambrish then asked whether we should create a blanket statement that the new parameters are preferred over the existing PAM4 parameters. Could we then avoid having to enumerate the new and old parameters in multiple locations? The group considered this but ultimately decided that it wasn't necessary. There weren't enough places in the text where we had to enumerate both the existing and new parameters to make it worthwhile to create new more general language. Randy said that we planned to add a sentence encouraging usage of the new parameters, but we hadn't yet drafted it. He said we should do it if we want model makers to transition to the new parameters. On page 278, in the Usage Rules for the existing parameter Modulation, Arpad had added a copy of the sentence stating that Modulation and Modulation_Levels shall not both be present. On page 280, in the Usage Rules for the PAM4_Mapping, Arpad had extended the list of conditions under which PAM4_Mapping is ignored to include when Modulation_Levels is declared. The group agreed with the extension of the existing language for now. Ambrish asked whether we should consider allowing the use of PAM4_Mapping with the new Modulation_Levels because there is nothing analogous to the PAM4_Mapping parameter in the new PAMn parameters. Arpad agreed that mapping is the one area where the new parameters are not a superset of the PAM4 parameters. Ambrish said someone who wants to specify Gray code will be inclined to use the old parameters instead of the new ones. Bob and Arpad said that if a model maker wants to specify Gray code they must use the old parameters. Arpad added a note to the draft stating that we should think about this issue later: Should we allow PAM4_Mapping to be used with the new modulation parameters? The group then returned to rewriting the introduction paragraphs in 10.7. Arpad added a new sentence extending the historical sequence and stating that IBIS 7.2 introduced (will introduce) PAMn support. Arpad and the group added sentences stating that the new parameters are preferred, and encoding is not specified with the new parameters because it is expected to be provided by the EDA tools. Ambrish noted that the sentences related to waveform data, transitions, etc., were generally applicable to NRZ, PAM4, and PAMn. Randy suggested that these sentences could be moved to page 223 in the section that defines the symbol_time parameter for AMI_Init() (and possibly also in the AMI_GetWave() section). - Curtis: Motion to adjourn. - Randy: Second. - Arpad: Thank you all for joining. AR: Arpad to send an updated draft 19 of BIRD213.1 to the ATM list. ------------- Next meeting: 29 March 2022 12:00pm PT ------------- IBIS Interconnect SPICE Wish List: 1) Simulator directives