IBIS Macromodel Task Group Meeting date: 31 May 2011 Members (asterisk for those attending): Agilent: Fangyi Rao * Radek Biernacki Ansoft: Chris Herrick Danil Kirsanov Ansys: Samuel Mertens * Dan Dvorscak Deepak Ramaswamy Jianhua Gu Curtis Clark Arrow Electronics: Ian Dodd Cadence Design Systems: Terry Jernberg * Ambrish Varma Celsionix: Kellee Crisafulli Cisco Systems: * Mike LaBonte Stephen Scearce Ashwin Vasudevan Ericsson: Anders Ekholm IBM: * Greg Edlund Intel: Michael Mirmak LSI Logic: Wenyi Jin Mentor Graphics: * John Angulo Vladimir Dmitriev-Zdorov Zhen Mu * Arpad Muranyi Micron Technology: Randy Wolff NetLogic Microsystems: Ryan Couts Nokia-Siemens Networks: * Eckhard Lenski Sigrity: Brad Brim Kumar Keshavan Ken Willis SiSoft: * Walter Katz Mike Steinberger * Todd Westerhoff Doug Burns Snowbush IP: Marcus Van Ierssel ST Micro: Syed Sadeghi Teraspeed Consulting Group: Scott McMorrow * Bob Ross TI: Casey Morrison Alfred Chong Vitesse Semiconductor: Eric Sweetman Xilinx: Mustansir Fanaswalla The meeting was lead by Arpad Muranyi ------------------------------------------------------------------------ Opens: - Arpad: No meeting next week due to IBIS summit - Will present an ATM update -------------------------- Call for patent disclosure: - None ------------- Review of ARs: - Arpad update Table Syntax Clarification BIRD: - Done - BIRD 132 Arpad update BNF BIRD draft: - Done ------------- New Discussion: Arpad: Various new documents are posted on the ATM website Walter showed the Add Boolean to BNF for IBIS AMI BIRD draft: - Arpad: The example has '' - Walter: Would it have quotes? - Arpad: No, it would be like a numeric literal - Radek: The "or" in section 8 does not sound right grammatically - Todd: Put "floating point number" before "ANSI ..." Walter showed the Corner Range BIRD draft: - Arpad: This has not changed since last week - There have been no comments - We could vote on this - Radek: What are the implications for EDA platforms? - Arpad: Type Corner was not spelled out clearly - This would end confusion - Walter: IBIS already has this concept - Arpad: Some tools have more than 3 combinations of IBIS data - Capacitance can be mixed with I-V strength - Walter: In AMI slow means slow - Radek: We should stay away from discussion other aspects of IBIS here - Todd: Users want to control typ/slow/fast - IBIS did not set a clear correlation with typ/min/max - AMI is intentionally different - Bob: The models have typ/min/max, the tools just mimic that - Arpad: No one says max V-T can be mixed with min I-V - Compensating buffers make it interesting - Walter: Models expect the EDA tool to pick slow for slow, etc. - Any other expectation should use List - Arpad: Please send comments by email Walter showed the AMI_parameters_out Clarification BIRD draft: - Arpad: There have been no comments - Walter: We should have notice the previous Friday of BIRDs to consider voting - I have emailed my comments - Arpad: Some of the new content is about memory allocation Arpad: Some examples in the spec have obviously wrong syntax - BIRD 127.1 will be submitted to Open Forum Walter showed the AMI Function Return Value Clarification BIRD draft: - Arpad: Ambrish suggested we should use this to send messages - It only exists for Init, but messages would be more from GetWave - Todd: "execution" should be changed to "configuration" - Arpad: Parameters can be used for GetWave feedback - Ambrish: We can vote to submit this - The vote passed by acclamation AR: Arpad submit AMI Function Return Value Clarification BIRD draft to Open Forum Walter showed the agenda: - Arpad: We should discuss status of the Out/InOut BIRD draft - Walter: We had a discussion on rewording it - Todd: It was about the use of "must not" - Arpad: I can change the wording about compliance - Walter: We should wait until after the summit - Bob: The spec should not get into too much detail Walter showed the Type and Format BIRD draft from Ambrish - Ambrish: There is a small grammatical error - Todd: How can Gaussian be Integer type? - Walter: Integer should not be allowed - Some floats at times can be integers, but they have wider range - Bob: It would have to be removed in the text too AR: Ambrish remove checkmark for DjRj,Gaussian as Integer type from draft BIRD Walter showed his AMI Time Domain Flow chart - Slide 1: - Walter: This is the whole AMI process in one picture - IBIS 5.0 has a Jitter param that could be any of these 4 jitter types - The tool can use any input to jitter the stimulus to the model - BIRD 123 added 3 parameters: - Tx_DCD was clarified - Tx_RjSj - Tx_Sj - Tx_Sj_Frequency - Dual-Dirac, DjRj and Table don't tell how to generate jitter mathematically - Tx_Sj gives the math - Ambrish: Are these maximum values for the model? - Walter: Mike Steinberger had a good email answering that - I will send it - Walter: Ref Clock parameters were added - Scott request clock_times from the CDR - An external clock actually drives the CDR - The model maker does not know it's jitter behavior - These parameters allow for modeling that jitter - Walter: Rx_Noise can be injected at the receive point - Greg: Can each parameter be associated with a physical source? - Walter: They might be too general - For the model developer it may be a budget - Slide 2: - Walter: These can only be accounted for the the EDA tool analysis - It can only generate an eye - It needs to know clock distribution to generate BER - Bob: Rx_Receiver_Sensitivity is not mentioned - Walter: That is not in BIRD 123 but should be in this picture Walter: Ken suggested I remove Samples_Per_Bit from BIRD 121 - It would not be good to have models not working at any sample rate - Walter suggested removing Samples_Per_Bit from BIRD 121 - No objection AR: Walter remove Samples_Per_Bit from BIRD 121 Meeting ended. ------------- Next meeting: 14 Jun 2011 12:00pm PT Next agenda: 1) Task list item discussions ------------- IBIS Interconnect SPICE Wish List: 1) Simulator directives