04/16/02 Minutes of IBIS Quality Committee Conference Call IBIS Quality Committee Members: Adam Tambone Adam.Tambone@fairchildsemi.com *Barry Katz bkatz@sisoft.com *Benjamin P Silva benjamin.p.silva@intel.com Bob Ross bob_ross@mentor.org *Eckhard Lenski Eckhard.Lenski@icn.siemens.de Eric Brock ebrock@sisoft.com Gregory R Edlund gedlund@us.ibm.com Hazem Hegazy hazem_hegazy@mentorg.com *John Figueroa jfigueroa@apple.com *Kevin Fisher kfisher@sisoft.com *Kim Helliwell kimgh@apple.com Lynne Green lgreen@cadence.com *Mike Labonte mlabonte@hhnetwrk.com Peter LaFlamme plaflamm@amcc.com *Robert Haller rhaller@cereva.com *Roy Leventhal Roy_Leventhal@3com.com Sherif Hammad sherif_hammad@mentorg.com *Todd Westerhoff twester@hhnetwk.com Tom Dagostino tom_dagostino@mentorg.com Everyone in attendance marked by * John Figueroa volunteered to explore the recording option on the teleconference. This could be beneficial if someone was unable to attend or wanted to review the conversation. He is going to report back on the options. Continued working through checklist... MODEL TYPE Each buffer type has minimum, maximum, un-related data. Un-related data leads to confusion and improper tool operation. Proposed was that we define the minimum and the maximum per I/O type and give example of un-related data (ie. output cells with Vih and Vil). There will be issues here due to the fact that each of us has a minumum/maximum. This is a level of detail we will get into later TEMPERATURE RANGE For informational purposes. 25C is rarely typical. 40C is more reasonable. Discussed need for multiple typical temperatures but beyond scope. Need to make sure range is correct/complete. Vendor needs to make reasonable guess at what typical operation conditions are. Use models for typical models are often to either get correlation in the lab or to determine operational marigins under typical conditions when broken at PVT extremes. CORNER MODELS Corner models should be included. There are exceptions (ie measured data). IV and VT curves should be generated under identical conditions (else VT/IV mismatch). Data sheets should bound IV curves. VOLTAGE RANGE Unfortunate that IBIS makes curves VCC relative -- can't crosscheck Voltage to IV. Can't change without re-gen of IV/VT curves. MIN/TYP/MAX Across all parameters these need to be complete/correct. C_comp What is the source? (simulation, measurement, guessed). Driver/receiver C_comp values can be different. Often have to make engineering judgement as to what is the proper value. As an example, in one application it may be perfectly acceptable to average driver/receiver C-Comp values (ie. 4pF vs 3pF -- use 3.5pF) but in others it may. Document what was done and show due diligence. STANDARD LOAD VREF, CREF, RREF, Vmeas should be in model for output cells. Should match data sheet (traceability to version) Some standard loads not supported in IBIS. If this is the case, should document in model. Should document in IBIS model if IBIS insufficiencies. IV CURVES Must look at actual IV curves. Visual inspection for Typ/Min/Max correctness. In some cases, results will be non-intuitvie (ie SS model stronger than FF due to process overcompensation). Note if non-intuitive. Some discussion as to whether a model should ever have MAX curves weaker than MIN curves -- food for thought. IBISCHK It was agreed that one of the outputs of our efforts would be to document problems with IBISCHK that prevent High-Quality IBIS models from passing. Roy was going to bring up making the IBIS golden parser open source at the next meeting.