================================================================================ IBIS INTERCONNECT TASK GROUP http://www.ibis.org/interconnect_wip/ Mailing list: ibis-interconnect@freelists.org Archives at http://www.freelists.org/archive/ibis-interconn/ ================================================================================ Attendees from April 29, 2020 Meeting (* means attended at least using audio) ANSYS Curtis Clark Intel Corp. Michael Mirmak* Ifiok Umoh Eric Edwards Keysight Technologies Radek Biernacki Mentor, A Siemens Business Arpad Muranyi* Micron Technology Justin Butterfield* Randy Wolff* SiSoft Walter Katz* Mike LaBonte Teraspeed Labs Bob Ross* Zuken USA Lance Wang* Michael Mirmak convened the meeting. No patents were declared. Justin Butterfield took minutes. Review of Minutes: - Michael called for review of the minutes from the April 22, 2020 meeting. Michael noted the word "caution" is misspelled. Walter Katz moved to approve the minutes with this correction. Randy Wolff seconded. The minutes were approved without objection. Review of ARs: - Randy to rerun the ibischk7 parser on the Interconnect examples with the caution flag enabled and report back. - Randy noted there is one caution for the ZQ pin, which has an NC model. The caution is related to the NC on a Pin_I/O. But, he was not sure of a better way to do this, as there is information in the package model for this pin. Michael commented he has seen similar issues. Bob Ross asked if this pin was included in the SPICE model. Randy replied this pin is included in the SPICE model, as they want to characterize the resistance. Bob commented, for SPICE, we need to include all the terminals. Randy stated this should not be an issue. - Bob to add the EMD Parts section in the EMD draft 7. - Bob sent out draft 7 with the EMD Parts section included. Opens: - None. EMD Draft Review: Michael asked if Bob could share the section where the EMD Parts text is placed. Bob shared the document, but it did not have the latest EMD Parts section. Michael asked where this will be placed. Randy noted it will replace the existing EMD Parts text. Bob will paste the EMD Parts section into EMD draft 8 and send it out [AR]. Michael asked if there were other changes, as mentioned in the email, and if the markup would show these changes. The email had noted changing "number of terminals" to "number_of_terminals". Randy commented we need more discussion about the examples on the use of number_of_terminals and signal_names. Bob noted one of the changes was for Example 3, where the interconnect model was split into three EMD models. The terminal numbers are now corrected, but a different subcircuit name should be used for these, since they have a different number of terminals. Randy asked about the syntax in Example 2. Bob added, in Example 2, we are adding a series resistor, but the same signal_name is used. This would be an error, since this would short across the resistor. Michael asked if the parser will check this. Bob replied, yes, this would be flagged as error, as the same terminal cannot be used more than once. Randy asked the reason for including the signal_name in the pin list, since in this case, the resistor makes the extended net. Bob noted the figure shows different signal_names, but this is not the case for this example. Randy commented, in the Example 3, the nets are different, but this is not the case for Example 2. Example 3 has the interconnect broke into separate subcircuit models. Michael asked how much checking will be done by the parser. Bob replied we need to discuss the rule, which will determine the parser behavior. Randy stated there are two cases. One with the resistor embedded, and one with the resistor brought out as a separate designator. It would be good to know how to label the signal_name for these cases. Walter noted the second column in the example is not correct. Michael asked if the parser will check the signal_name should be different unless it is a rail. Walter replied this is not the case. Randy commented the issue is how to label the signal_name, and how do we know the A07r exists and if this needs to be listed. Walter noted one side of the resistor would have the A07r signal_name. Michael was concerned that the parser would need to understand the path from inside the netlist to check this. Walter stated this is known from the the signal_name. If the resistor is inside the interconnect model, then the resistor net does not need to be defined in the pin list. Michael asked where the A07r should be defined. Walter stated there are separate nets for A07r and A07. Bob commented we do not have an extended net syntax. Walter stated the EDA tool will have to determine the extended nets. Michael asked if the parser will check each signal_name has at least two pin names, and if we are checking for unconnected components. Walter noted that this could be possible for NCs. Randy proposed for the example to use A07r as the signal_name. Bob agreed. Randy asked if the EDA tool can determine how to set up a simulation for this example. Walter replied the EDA tool will look at the pin list to determine the connections. Michael asked if there would be any way to distinguish a pin with only one connection. Walter stated this would be possible. Randy asked, in the case of a resistor package, how the EDA tool would know which pins are connected. Walter stated the EDA tool would have to determine the connections. Lance Wang asked if the resistor can be defined inside the subcircuit. Randy noted this is the case for Example 1. Bob commented this would only have A0 as the signal_name. Lance asked if the syntax could be simplified to require the model maker to put the series resistor in the subcircuit. Walter replied it is much easier to extract separate nets for the extended nets and add the series resistor externally. The tool does not know which resistor created the extended nets. Randy asked if this is a situation where we would want to make a syntax to help the EDA tool. Walter stated the EDA tool would need some help to determine the connections and the tool will have to deal with that, but we do not have a way to define how resistor packages are defined. Bob asked if "Ohm" can be changed to "ohm" in lower case. Michael and Walter agreed this is preferred. Next Meeting: The next meeting will be May 6. Randy moved to adjourn. Walter seconded. The meeting adjourned without objection. ================================================================================ Bin List: EMD Comments to be Resolved: (See BIRD202.1 tracking spreadsheet) IBIS-ISS Parser: - IBIS-ISS parser scope document