================================================================================ 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 March 20, 2019 Meeting (* means attended at least using audio) ANSYS Curtis Clark Cadence Design Systems Bradley Brim* Intel Corp. Michael Mirmak* 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* 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 June 27, 2018 meeting. Arpad Muranyi moved to approve the minutes. Brad Brim seconded. The minutes were approved without objection. Review of ARs: - None. Opens: Mike LaBonte stated we could discuss the decoupling model proposal. Bob Ross noted that per the discussion in the ATM group Walter Katz is generating an email about the proposal to be reviewed. Mike suggested to wait for the email to be drafted to have further discussions. Michael asked if there is any work for the Interconnect group on the proposal. Walter noted that he is working on the email with some questions and showing the same decoupling model implementation in BIRD189 syntax. He plans to review this email in the next ATM meeting. Arpad noted he preferred to keep the decoupling model discussion in the ATM group, so that the Interconnect Group can focus on EMD. Arpad also asked Walter to only send the email for review to the people at the last ATM meeting to limit the distribution. Walter agreed. Touchstone Enhancements: Michael noted the last changes to Touchstone were in 2011 and that some of of the TSIRDs on the website have been approved and not yet added to a new Touchstone version. He asked if we want to work on a new Touchstone version and if there is still an industry need to include these features. Brad suggested we could also add a transfer function format. Walter suggested that the transfer function could be a new format separate from Touchstone. Bob thought that this could be done later. Walter noted that the main reason for Touchstone 2.0 was to add reference impedance per port capability. Brad noted that Cadence does support Touchstone 2.0, and he does see requests for Touchstone 2.0 model support. Arpad stated that Mentor does support Touchstone 2.0 as well, but asked if Touchstone 2.0 models are widely supported by extraction tools. Walter suggested that we should start from Touchstone 1 and make enhancements as a new Touchstone 1.x, since Touchstone 2.0 is not widely adopted in the industry. Brad noted it would not make sense to invest in a Touchstone 1.2 while not making the same changes to a Touchstone 2.1. Michael noted that Intel does support and generate Touchstone 2.0 models in some cases. Arpad noted that SI/PI co-simulation is enabled with the per port reference feature, and Mentor's tools are supporting this. He asked what mistake we made with Touchstone 2.0 and if we could simply add features back into a new Touchstone version to fix the issue. Walter replied that the mistake we made is that Touchstone 2.0 broke existing Touchstone parsers, and there are so many Touchstone parsers. The only benefit to Touchstone 2.0 was for people doing PI simulation and the need for per port referencing. Brad noted that the extraction tools that write the models do not write Touchstone 2.0 and that Cadence has to support proprietary Touchstone formats. Walter commented that, if we defined Touchstone with records to define the per port reference, old parsers would still work. The problem with Touchstone 2.0 is that it added keywords that will break old parsers. Bob commented that EDA companies can buy the Touchstone 2.0 parser for $1000, so cost should not be an issue. Arpad asked if the problem is that people don't upgrade their parsers. Walter replied that, if they need the per port reference feature only, then they would need to change their parser. Bob asked who Walter is referring to. Walter replied he is talking about the users who have in-house custom tools that have a Touchstone parser built in. Michael suggested to do some data gathering to help answer this question. He asked what is the best way to answer these questions. Arpad suggested to send an email to the SI list with some questions. Walter agreed with using SI list to get the best coverage. Michael asked if there is someone who can draft an email to the list. Brad suggested that who ever sends the email needs to represent the IBIS Open Forum. He also noted that most Touchstone files are written by the EDA and measurement tools. Walter stated the critical question is the usage in custom software. Bob disagreed and asked how many companies would still have their own analysis tools. Brad agreed with Walter and noted that if industry leaders write Touchstone 2.0 files others will follow. Michael asked who can write the draft email survey to the SI list. Brad stated he can help review the email, but he is retiring next month. Arpad suggested Michael should draft the email. Mike noted he can send the email on behalf of the IBIS Open Forum. Michael and Mike will draft an email survey to the SI list on the usage of Touchstone 2.0 models [AR]. EMD Status: Michael asked if Walter has a presentation on EMD that he can go over. Walter replied that he can present on the EMD draft he currently has. IBIS-ISS Parser: Michael commented he would still like to discuss the idea of an IBIS-ISS parser. Mike stated that they discussed this in the Quality group meeting, but only he and Bob are active in that meeting. Arpad noted that this is not a high priority, as every EDA tool will have to parse these files when they read them into their spice engine. Bob noted there are two issues. The first is if we want to have the parser check IBIS-ISS syntax as it is now. And, the other issue is if the parser should be attached to the ibischk parser or if there is a separate IBIS-ISS mode that can be run independently. Next Meeting: The next meeting will be March 27. Mike moved to adjourn. Bob seconded. The meeting adjourned without objection.