IBIS Macromodel Task Group Meeting date: 1 September 2009 Members (asterisk for those attending): Adge Hawes, IBM * Ambrish Varma, Cadence Design Systems Anders Ekholm, Ericsson * Arpad Muranyi, Mentor Graphics Corp. Barry Katz, SiSoft * Bob Ross, Teraspeed Consulting Group Brad Brim, Sigrity Brad Griffin, Cadence Design Systems Chris McGrath, Synopsys David Banas, Xilinx Deepak Ramaswany, Ansoft Donald Telian, consultant Doug White, Cisco Systems * Eckhard Lenski, Nokia-Siemens Networks Essaid Bensoudane, ST Microelectronics Fangyi Rao, Agilent Ganesh Narayanaswamy, ST Micro Gang Kang, Sigrity Hemant Shah, Cadence Design Systems Ian Dodd, consultant Jerry Chuang, Xilinx Joe Abler, IBM * John Angulo, Mentor Graphics John Shields, Mentor Graphics Ken Willis, Cadence Design Systems * Kumar Keshavan, Sigrity Lance Wang, Cadence Design Systems Luis Boluna, Cisco Systems Michael Mirmak, Intel Corp. * Mike LaBonte, Cisco Systems Mike Steinberger, SiSoft Mustansir Fanaswalla, Xilinx Patrick O'Halloran, Tiburon Design Automation Paul Fernando, NCSU Pavani Jella, TI Radek Biernacki, Agilent (EESof) * Randy Wolff, Micron Technology Ray Komow, Cadence Design Systems Richard Mellitz, Intel Richard Ward, Texas Instruments Samuel Mertens, Ansoft Sam Chitwood, Sigrity Sanjeev Gupta, Agilent Shangli Wu, Cadence Design Systems Sid Singh, Extreme Networks Stephen Scearce, Cisco Systems Steve Pytel, Ansoft Syed Huq, Cisco Systems Syed Sadeghi, ST Micro Ted Mido, Synopsys * Terry Jernberg, Cadence Design Systems * Todd Westerhoff, SiSoft Vladimir Dmitriev-Zdorov Vikas Gupta, Xilinx Vuk Borich, Agilent * Walter Katz, SiSoft * Zhen Mu, Mentor Graphics ------------------------------------------------------------------------ Opens: Mike L will be unable to take notes next week - Randy can fill in -------------------------- Call for patent disclosure: - No one declared a patent. ------------- Review of ARs: - Arpad send notice that his and Kumar's documents are posted - Done - Arpad send updated IBIS-ISS presentation to Walter - TBD - Mike change web page to make items easily linkable - TBD - Arpad write a BIRD to clarify time period accuracy requirements - TBD - Todd: Write IBIS s-param BIRD - No update - 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: Kumar showed his document "IBIS ATM: txgetwave" - Slide 3: - There is a big difference between a bit generator and an equalizer - Slide 4: - The bit waveform can have all kinds of distortion - Slide 5: - Tx_GetWave can have any continuous waveform if the model is LTI - Slide 7: skipped - Slide 8: - The analog front end is separate from the bit generator - Slide 9: - 0 to 1 signals are easier to work with than -0.5 to 0.5 - Walter: Agree, but with 2 exceptions: - Slide 7: should be "or is time varying", not "and is time varying" - Slide 9: Input waveform must go from -0.5 to 0.5 Arpad showed his document "Comparing Walter's and Kumar's IBIS-AMI Proposals" - Slide 2: - Do not believe there is an actual disagreement - Slide 3: - In Walter's example the 75% DCD makes it non-linear - Todd: The question is about where the DCD is introduced - Walter: If TX EQ is truly time-invariant and amplitude-invariant, then DCD does not make it LTI - A peaking filter DCD does not make it non-LTI - Kumar's presentation is precise and correct - But an FIR filter does not fit the category - Kumar: Disagree with that last point - Arpad: Kumar's idea is based on bit pattern independence - Slide 5: - Kumar's assertion agrees with Walter's - Walter: We will not be able to settle that this week - Kumar: It does not depend on the bit stream - I used a different symbol (tau) in my equation - Todd: Uppercase letters are F domain, lowercase are T domain - Kumar is defining new terms when we have already agreed on terminology - Arpad: The only discrepancy is that Walter included convolution with the channel response in Tx_GetWave - Kumar did not - Slide 8: - Walter: There are technical issues to resolve Walter showed a composed email: - It contains the same points Walter made today - It also recommends changes to IBIS AMI section 2.3 - A new flow should be described - We can discuss this next week - Ambrish: After changing the flow, if the TX_GetWave model is LTI, would the flow remain the same as what we passed in IBIS 5.0? - After making the change, we will distinguish between a TX model and an RX model as far as the actual models are concerned. - In the current spec, there is no distinction. - Todd: Is Tx_GetWave output at TX pad or RX pad? - Walter: We agree it should be TX pad, which goes into the channel - Arpad: The EDA tool performs the convolution between TX and RX AR: Walter send his email proposing AMI flow changes Walter showed document "Init should return just buffer equalization": - Slide 4: - Different tools are used to make RX Peaking Filters and DFE EQ - Matlab, SPICE, etc. - Slide 6: - How to generate TX_GetWave input: - Sample data - SPICE - AMS - Matlab - We should have on option for RX_Init to return just hREI - A reserved parameter would flag this - Arpad: After RX_Init we have Tx_GetWave - This is getting in the way - We should go through both TX then both RX - Walter: There is no need for Tx_GetWave for LTI TX - Existence of TX_GetWave implies non-LTI - Kumar: If you use Tx_GetWave you don't have Tx_Init - Todd: The model should just return the impulse response - EDA tools have a hard time recovering the impulse response AR: Walter send Init_Return_Just Buffer_Equalization presentation to Mike L AR: Mike L post presentations from Walter to web page Walter suggested Mike L send minutes to Arpad, Kumar and Walter for preview Next meeting: 15 Sep 2009 12:00pm PT -------- IBIS Interconnect SPICE Wish List: 1) Simulator directives