Minutes, IBIS Quality Committee 17 November 2009 11-12 AM EST (8-9 AM PST) ROLL CALL Adam Tambone * Anders Ekholm, Ericsson Barry Katz, SiSoft Benny Lazer Benjamin P Silva Bob Cox, Micron * Bob Ross, Teraspeed Consulting Group Brian Arsenault David Banas, Xilinx * Eckhard Lenski, Nokia Siemens Networks Eric Brock Guan Tao, Huawei Technologies Gregory R Edlund Hazem Hegazy Huang Chunxing, Huawei Technologies John Figueroa John Angulo, Mentor Graphics Katja Koller, Nokia Siemens Networks Kevin Fisher Kim Helliwell, LSI Logic * Lance Wang, IOMethodology Lijun, Huawei Lynne Green, Green Streak Programs * Mike LaBonte, Cisco Systems Mike Mayer, SiSoft Moshiul Haque, Micron Technology Muniswarareddy Vorugu, ARM Ltd Pavani Jella, TI Peter LaFlamme Randy Wolff, Micron Technology Radovan Vuletic, Qimonda Robert Haller, Enterasys Roy Leventhal, Leventhal Design & Communications Sherif Hammad, Mentor Graphics Todd Westerhoff, SiSoft Tom Dagostino, Teraspeed Consulting Group Kazuyoshi Shoji, Hitachi Sadahiro Nonoyama Everyone in attendance marked by * NOTE: "AR" = Action Required. -----------------------MINUTES --------------------------- Mike LaBonte conducted the meeting. Call for opens and patent disclosures: - No one declared a patent. - Mike has an ongoing meeting conflict - Would like to avoid running over 1 hour to rejoin other meeting AR Review: - Mike upload new IQ spec and fix web page to link it - Moshiul rework IQ checklist for IQ 2.0 - Lance try to find IBIS files with [Diff Pin] Tdelay out of order - Talked to TI - Unable to release any files - Mike: This is enough to prove that typ/min/max swizzling is real New items: Mike showed the IQ web page: - It now has links directly to ver1 and ver2 docs - Bob: Would anyone be able to operate the web tools that create this? - Mike: It isn't that easy, takes a series of steps to add something - quality and macromodel use the same Perl software - There is a README in the macromodel directory We looked at the IBIS Accuracy Handbook: - Mike: Test loads should be discussed in the context of Golden Waveforms - Bob: This is written with legacy CMOS in mind - It never addressed differential or ECL - Implies bipolar in some places - Mike: Section 3 is about how to make IBIS models - This should be in in the cookbook - Section 4 is about correlation - Table 3: Correlation Levels - Mike: Known typ/min/max samples can be provided by SPICE - Bob: Not a fan of just numbers - Like to see the waveforms too - Mike: Not inclined to have correlation level "buckets" - Anders: It doesn't really add anything - Bob: Huawei likes to correlate against real boards - Anders: They also adjusted the buffers on the board - Not sure how they would know what corner they are in - Eckhard: They may be comparing against the IBIS models - Lance: They modify the IBIS models to match their measurements - You would have to be sure your measurements are right - 4.3 Curve Overlay Metric - The FOM algorithm is a simple unweighted error summation - Bob: Randy Wolffe has a presentation last year - Used I-V strength scaling to improve correlation - Then update V-T curves - Bob: Tom Dagostino may have a presentation on sampling issues - Measurement does not show wild swings - Anders: You need multiple loads to get it right - JEITA does 12 loads - Mike: So far this has 3 parts: 1) How to extract IBIS - Should drop this, it belongs in the Cookbook 2) Test loads - We should give direction on this - Maybe people will start to generate Golden Waveforms, given direction 3) Correlation measures - David and Roy got us going down the feature-selective measurement path - Bob: We should just leave the Accuracy Handbook behind - Bob: The FOM is a voltage domain measure, not time domain AR: Mike notify ibis and ibis-quality lists about correlation discussions Lance: There is a new "IBIS Healing" tool - There were problems running it - Bob: You have to expand the page and hit Next - Guy DeBurgh found this tool - It came from a professor in Italy affiliated with UMR - Mike: We might want to look at it Meeting ended at 12:11 PM Eastern Time.