IBIS Macromodel Task Group Meeting date: 06 October 2015 Members (asterisk for those attending): ANSYS: * Dan Dvorscak * Curtis Clark Avago (LSI) Xingdong Dai * Bob Miller Cadence Design Systems: * Ambrish Varma Brad Brim Kumar Keshavan Ken Willis eASIC * David Banas Marc Kowalski Ericsson: Anders Ekholm IBM Steve Parker Intel: * Michael Mirmak Keysight Technologies: Fangyi Rao * Radek Biernacki Maxim Integrated Products: Hassan Rafat Mentor Graphics: * John Angulo * Arpad Muranyi Micron Technology: * Randy Wolff Justin Butterfield QLogic Corp. James Zhou Andy Joy SiSoft: * Walter Katz Todd Westerhoff * Mike LaBonte Synopsys Rita Horner Teraspeed Consulting Group: Scott McMorrow Teraspeed Labs: * Bob Ross TI: Alfred Chong (Note: Agilent has changed to Keysight) The meeting was led by Arpad Muranyi. ------------------------------------------------------------------------ Opens: - None. -------------------------- Call for patent disclosure: - None. ------------- Review of ARs: - None. ------------------------- Review of Meeting Minutes: - Arpad: Does anyone have any comments or corrections? [none] - Radek: Motion to approve the minutes. - Mike L: Second. - Arpad: Anyone opposed? [none] ------------- New Discussion: Item #6: Info, Out, InOut BIRD draft. - Arpad: Last week we decided: - No additional "severity" indicator parameter. - The list of special Model Specific parameters would have to include all of the parameters requiring special handling, even those parameters that are for "above and beyond" features not required for the basic simulation flow. - Arpad: Walter has raised a new question regarding the analog modeling. - Walter suggested adding a new subparameter under [Algorithmic Model] in the IBIS file. - I agree with his suggestion, but we need to discuss it. [below] - Arpad: [sharing a draft of the BIRD] - Today I want to work on the final verbiage. - Discussion: Parser instructions/requirements added to the "Any Other Background Information:" section of the BIRD draft. There was general agreement that the parser should issue an error if the new Special_Param_Names parameter contained the name of a parameter that was not found in the Model Specific parameters section. There was debate over whether the parser should issue a warning or a note if the Special_Param_Names exists at all. Ambrish suggested that it should issue a warning to make sure the user was aware of potential portability issues. Walter disagreed and said that we didn't want to punish model makers for using the new parameter. Mike L. noted that "forcing" model makers to use this feature and then issuing parser warnings might be ill-advised. He noted that it was really up to the EDA tool, not the parser, to process the information from Special_Param_Names and present the user with the correct compatibility information. Bob R. was asked to define the difference between "note" and "caution" in the parser output. Bob said that notes are always printed, and caution messages must be enabled by setting a command line flag. He noted that their actual "levels" were similar. The thing being flagged in either case was not a technical requirement. Radek stated that caution would probably be a better choice than note, were it not for the fact that notes were always displayed and cautions were not. But he said he was fine with using note in this case. Arpad and Mike L. suggested that the details could be ironed out in the IBIS quality meeting, but Walter was firm that he wanted the text of the BIRD draft to state that a note would be issued and not a warning. Arpad made the change and noted that we could return to this discussion later. Walter suggested that we had to complete the text of the rest of the BIRD to fully work out exactly what happens when a parameter is listed in this new Special_Param_Names. - Discussion: Final wording of the "Definition" and "Usage Rules" for the new Special_Param_Names. Arpad presented several proposals for defining the parameter. The group worked on the text real-time and arrived at the following: Definition - This reserved parameter identifies, by name, all Model_Specific parameters that require EDA tools to perform special handling that is not described in the IBIS specification. Usage Rules - If the .ami file in which this reserved parameter appears contains any Model_Specific parameters associated with special operations that the model expects the EDA tool to perform beyond what is described by the IBIS specification, the name of all such Model_Specific parameters must be listed in this reserved parameter. - Bob R - I think the title of the BIRD should now officially be changed. - Arpad - [Changed title: New IBIS-AMI Reserved Parameter Special_Param_Names.] - I will clean up this draft based on these discussions. - Walter, would you like to discuss your analog model issue? - Walter: [sharing email on the topic] - Issue: IBIS 5.1, and even IBIS 6.0, is not going to describe the analog models with sufficient accuracy. - IBIS 6.0 now allows us to use s-parameters for AMI modeling, but there are still going to be models that require proprietary simulators. Many of our tools have the ability for models to point to other simulators, whether they do it in the IBIS file or somewhere else. - But that's an IBIS [Model] that does not contain an adequate description of the analog behavior of the buffer. - Shouldn't we have some method for the model maker to advertise that to the user? That is, to say, "you can't just take the I/V and V(t) curves and get a model with sufficient accuracy to do your job." - I propose adding a new IBIS subparameter, perhaps under [Algorithmic Model], that would advertise to the user that the analog [Model] is insufficient. - Proposed name: IBIS_Analog_Model_Incomplete - Could be true, false, etc., we can work out the exact syntax. - Tell the model maker community that when you have a [Model] that's insufficient you should advertise it. - Discussion: Bob R. said that a [Model] that needed this subparameter was effectively dead and shouldn't be released, except that there must be some special parameters that work in somebody's tool. Walter concurred and said this was often done with special keywords hidden behind comment characters, because the parser has no way of handling "special" keywords in the .ibs file the way it does for dealing with Model Specific .ami parameters. Walter wondered if we might not want to add something equivalent to Model Specific parameters to the .ibs file. Arpad and Bob R. both suggested that this proposal should be written up separately from the Special_Param_Names BIRD, despite the fact that they dealt with somewhat similar issues. Michael M. noted that a similar new subparameter could reasonably be defined for [Package Model], and any other area where the specification might obligate the model maker to provide some data that would not be entirely useful or complete. - Arpad: I will work on cleaning up a new draft of the Special_Param_Names BIRD. - We can work on this analog modeling proposal under a separate BIRD. - Thank you all for joining. AR: Arpad to clean up draft 14 of the Special_Param_Names BIRD and send it to Mike L. for posting to the ATM website. ------------- Next meeting: 13 October 2015 12:00pm PT ------------- IBIS Interconnect SPICE Wish List: 1) Simulator directives