Minutes from four ibis-quality meetings are attached. Reminder: there is no ibis-quality meeting this Tuesday. Mike
Minutes, IBIS Quality Committee 16 March 2010 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 Tim Coyle, Signal Consulting Group 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: The agenda email had line wrap issues AR Review: - Bob consider creating ibis/quail directory and Perl script - This is not a high priority - The directory quail_wip exists, the first QUAIL has been uploaded - It is not linked to the main page yet - We may change our minds - This process is not for the Open Forum until we have a new IQ release New items: Discussion of IBIS web pages (off-topic): - Bob: There are lingering issues and we need to resolve them soon - eigroup.org pages are a problem - Some changes made on eda.org are having trouble too - Moshiul: Links are broken, the top page links to itself - Bob: We need to have everything in subdirectories with strict ownership - Bob: We should make sure that all admin accounts have redundant coverage - Mike: On eda.org we have the 'ibis' group permission - But some files lack it - Bob: We need double coverage on freelists.org Develop outline for IBIS Correlation document: - Mike: Unable to find outline document created with Pavani's help - Will keep looking Mike showed the quail_wip directory created by Bob: - It has only the QUAIL_2.0-1.txt file, no HTML - Bob: It is not listed in directory.html - We may put in structure similar to the BIRD or BUG structure Agenda for next week: - Bob: David Banas' version of FSV is perfectly valid - It should be included - A good model can show bad results - Moshiul: Can send some some ideas - Mike: Does the NSN IBIS website have correlation info? - Eckhard: No Next meeting will be Mar 23 Meeting ended at 12:05 PM Eastern Time.
Minutes, IBIS Quality Committee 23 March 2010 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 Tim Coyle, Signal Consulting Group 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 IBIS related patent disclosures: - Bob: The role of this committee is to produce specifications for approval - The Open Forum must approve to be official IBIS - They are unlikely to become ANSI standards - Not sure if the cookbook and handbook are official - Mike: The outline document created with Pavani's help has been found - No one declared a patent. AR Review: - None New items: Mike showed previous documents related to correlation: - David's FSV paper - Mike: It has 2 voltage and time domains - Lance: For Vol/Voh it should compare what IBIS got against SPICE - It overlooks the shape of the waveform near the switching - Bruce Archambault showed us a method that looks at the whole wave - Mike: There may be a place for both methods - Moshiul: It still gives an idea Mike showed Guan Tao's Measurement Correlation presentation: - Lance: They check real board measurements against IBIS - If it correlates the IBIS model becomes golden - Mike: This tests far more than the IBIS model - Page 10: - Mike: This is similar to David's recommendations - Moshiul: What is "High level valid pulse?" - Mike: That may be the time above Vih - That is a good test - Page 9: - Moshiul: There are places where the waveform differences are large - But the correlation by what they measure still is good - This is like an envelope metric - Mike showed the IBIS Correlation outline: - Mike added F/V/T domains for FSV - Mike: Driver I/V strength may not be important - Moshiul: We check this frequently - We compare against the datasheet if it has it - Mike: ODT is important - Moshiul: We mostly check the ODT strength at Vtt - Checking the whole range would be good - Mike: Do we need to correlate clamp I/V curves? - Eckhard: Clamps need to be accurate for PCI bus - Mike: Receiver Output correlation might be going too far - Moshiul: agree - Mike: We should be able to describe most fixtures with [Test Load] - If not, maybe a BIRD is needed to extend [Test Load] - Moshiul: Sometimes test boards are used - [Test Load] may not be enough - Mike: EMD will be worked on this year - Moshiul: Will that replace ICM? - Bob: The ATM group's first priority is AMI AR: Mike post updated outline of IBIS Correlation Handbook Next meeting will be Mar 30 Meeting ended at 12:20 PM Eastern Time.
Minutes, IBIS Quality Committee 30 March 2010 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 Tim Coyle, Signal Consulting Group 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 IBIS related patent disclosures: - No one declared a patent. AR Review: - Mike post outline - Done, but no announcement New items: Bob showed the IBIS Correlation Outline and made updates: - Mike: Does any part of this not make sense? - Anders: Why is there not a Simulation section? - Mike changed Measurement to "Measurement and Simulation" - Anders: The correlation has to check the same node on both sides - Moshiul: The correlation setup belongs in section 3 - Anders: Why do we have Receiver Output correlation? - We deleted that - Anders: Input correlation is not that same as others - Mike: Is die capacitance comparison really correlation? - Moshiul: This is effective capacitance as seen at the pin - Anders: What will be our measurement points for that? - Mike: Maybe we could drop Effective Capacitance correlation - Anders and Moshiul wanted to keep it Mike: Should correlation be against actual data in an IBIS file? - Maybe it should always be against IBIS simulation, not IBIS data - Moshiul: Some IBIS data should be checked directly - Mike: I/V curves are extracted and recombined by simulators - Doing that and checking correlation could be worthwhile - Anders: This can be simulator dependent - What is the scope of this? - Anders: We should correlate the IBIS data, not simulators - Bob: We should not make it a big project - The IBIS model is not the only variable - Measuring it through parasitics complicates things - Mike: - There are two parts: - What to correlate - How to measure correlation - We should focus on the latter Bob: We should keep the input section - The effect of inputs on reflection is significant Anders: Who would be using this document? - Mike: IC vendors - Moshiul: Mostly - Mike: Maybe our document should have 2 sections: - sim-sim correlation - sim-data correlation AR: Bob send updated IBIS Correlation outline to Mike AR: Mike post updated IBIS Correlation outline to IQ web Next meeting will be Apr 12 Meeting ended at 12:14 PM Eastern Time.
Minutes, IBIS Quality Committee 06 April 2010 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 Tim Coyle, Signal Consulting Group 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 IBIS related patent disclosures: - No one declared a patent. - Eckhard mentioned an email from Colin Warwick about crowd-sourcing - Lance: It will not work - Companies will not assign people to update the models - Mike: Companies want us going to their web page - Bob: We should not take up this project - If someone else does it that's fine - Eckhard: It appeared only on IBIS AR Review: - Bob send updated IBIS Correlation outline to Mike - Done - Mike post updated IBIS Correlation outline to IQ web - Done New items: Eckhard: Should we have differential input correlation? - Mike: Do we want to check IBIS data against the data it was extracted from? - Lance: Accuracy goes down when a different common mode voltage is used - Mike: The test we correlate to should be "real world" - Lance: Bob mentioned the problem of non-linear loads some time ago - Mike: Moshiul said that he likes checking ODT - Eckhard: Moshiul was talking about driver I/V Mike: What do we want for top level organization? - Our outline may have too much Measurement before Correlation - Bob: This looks like the Accuracy Handbook - Mike: The correlation should be against an IBIS simulation - It should not be a double-check of the extraction - Bob: FSV is the really new part here - There are special issues with differential too - Mike: We should discuss Correlation first - Bob: Agree Review of sections under Correlation - Mike: Should FSV be split into 2 categories? - Frequency domain FSV - Voltage and time domain FSV - Bob: FSV has an IEEE standard for the frequency domain approach - Mike: The frequency domain approach is nice - It is easy to do - Eyeballing voltages usually only compares low frequency content Mike: Correlation ii through v could be moved to Measurement and Simulation - Section 3 could be seen as "How to Check Correlation" - Section 4 could be "Where to check Correlation" Mike: Comparing bench measured I/V against raw IBIS data is useful - Bob: Moshiul does this with scaling Next meeting will be Apr 13 Meeting ended at 12:15 PM Eastern Time.