IBIS Macromodel Task Group Meeting date: 27 Nov 2007 Members (asterisk for those attending): * Ambrish Varma, Cadence Design Systems * Arpad Muranyi, Mentor Graphics Corp. Barry Katz, SiSoft * Bob Ross, Teraspeed Consulting Group David Banas, Xilinx Donald Telian, consultant Doug White, Cisco Systems Essaid Bensoudane, ST Microelectronics Ganesh Narayanaswamy, ST Micro Hemant Shah, Cadence Design Systems * Ian Dodd Joe Abler, IBM * John Angulo, Mentor Graphics John Shields, Mentor Graphics Ken Willis, Cadence Design Systems * Kumar, Cadence Design Systems Lance Wang, Cadence Design Systems Luis Boluna, Cisco * Michael Mirmak, Intel Corp. * Mike LaBonte, Cisco Systems Mike Steinberger, SiSoft Patrick O'Halloran, Tiburon Design Automation Paul Fernando, NCSU * Radek Biernacki, Agilent (EESof) * Randy Wolff, Micron Technology Ray Comeau, Cadence Design Systems Richard Ward, Texas Instruments Sanjeev Gupta, Agilent Shangli Wu, Cadence Stephen Scearce, Cisco Systems Syed Huq, Cisco Systems Syed Sadeghi, ST Micro * Todd Westerhoff, SiSoft * Terry Jernberg Vikas Gupta, Xilinx Vuk Borich, Agilent * Walter Katz, SiSoft -------------------------- Call for patent disclosure: No one declared a patent. ----- Opens: - Arpad asked if a Dec 4 meeting with the interconn group is OK - Michael M.: Should be a good time for a joint meeting ------------- Review of ARs: - Walter coordinate completion of true diff BIRD draft - In progress - Michael M. draft Eye keywords for true diff BIRD - Will ask questions this meeting - Randy draft derating keywords for true diff BIRD - TBD - Arpad: Write parameter passing syntax proposal (BIRD draft) for *-AMS models in IBIS that is consistent with the parameter passing syntax of the AMI models - TBD - TBD: Propose a parameter passing syntax for the SPICE - [External ...] also? - TBD - Arpad: Review the documentation (annotation) in the macro libraries. - Deferred until a demand arises or we have nothing else to do ------------- New Discussion: - ICM-ATM connectivity - There will be a joint ICM/ATM meeting in the next ATM time slot - Walter presented his draft true diff pair BIRD: - [Model Selector] has been removed to simplify the BIRD text - Mostly described by example - Diff_receiver_threshold subparameter removed from [Model] - [Diff Model] section: - Added single_ended_model subparameter - [External Model] is a question mark here - Michael M.: Do we need single_ended_model inside Model_type Pseudo_diff? - [Transmit Thresholds] added - Could have one tdelay typ/min/max keyword or 3 keywords - Michael M.: Will we use these enough? - Todd: Never seen this used - Arpad: True diff should have tdelay built in - Todd: Does tdelay double-count something that is already in timing spec? - Kumar: It depends on where the capacitance is - Arpad: tdelay is intended to control inverse pin launch time - More discussion of whether tdelay is still used ... - Bob: This is hardly used, we should not advance it - We decided not to have tdelay in [Diff Model] - Walter: [Diff Pin Model] takes precedence over [Diff Pin] - Bob: Prefer not to have [Diff Pin Model] - Walter: True diff and pseudo diff differ only in current flow between pins - Arpad: Thought we would get rid of kludgy true diff mechanism using [Diff pin] calling [External Model] with these new keywords - Order of precedence rules may not be needed - It should be an error if [Receiver Thresholds] found in [Model] is "called" by [Diff Pin Model] - Measurements are appropriate in [Diff Model] whether for true or pseudo - Added Vdiff to [Receiver Thresholds] - Mike L.: Concerned about overloading [Receiver Thresholds] - Could change to [Diff Receiver Thresholds], but not important now - Todd: Mike S. proposed derating keywords, do we need those? - Arpad: These would be part of [Diff Model], not separate - Kumar: It is not clear which location these apply to - Michael M.: The old Si_location and Timing_location subparameters are not used much - Bob: There is no "core" value for these, only Die and Pin - Michael M.: AMI can be used to get the "latch access" location - Todd: Some may not understand the derivation of their timing data - [Algorithmic Model] and [External Model] must be in [Diff Model] if called by [Diff Pin Model] - Arpad: Would like to have this content in official BIRD format - Walter: Can't have single-ended model data in [Model] with [External Model] - Arpad: [External Model] takes precedence, but other data is not prohibited - Kumar: We can make use of all data types - Walter: Then tools would allow the user to choose which type of model to use - Michael M.: This sounds like an alternative to [Model Selector] - Arpad: Should be able to meet Dec 4, 11, 18, Jan 8 - Review of this BIRD could be Dec 18 - Bob: Should be able to have [Model] and [Diff Model] of same name - Walter: This is terribly confusing in [Model Selector] - Bob: We should separate measurement criteria from output behavior - Bob: This is a complex spider web of rules and scoping - Mike L.: We should have a parser code developer look at this - Walter: All new keywords could go into regular single-ended [Model] - Bob: The syntax has advantages, but does it warrant the complexity? - Bob: P & N [Diff Model]s might be found with different data - Walter: Only the high side is used - Michael M: Should we work around old keywords to get new features or clean up old keywords? - Walter: These changes are needed today, and can be done quickly - Restructuring IBIS is a big project - Arpad: We can clean up incrementally - Walter: How long will it take? - Bob: 6 months to approve BIRD, 18 months for EDA tools to implement - Walter: EDA vendors already have structures to get these models - [Diff Pin Model] is not really needed - Extracting parameters from high side [Model] will suffice - Bob: Creating new keywords to replace old keywords causes problems - Walter: We need a more compelling reason to do this - Michael M.: The EDA vendors should discuss this AR: EDA vendors meet to decide differential BIRD direction Next meeting: 04 Dec 2007 12:00pm PT