IBIS Macromodel Task Group Meeting date: 17 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: - Mike L: Mark Westerhoff has joined as a guest - Also Mike can not attend next week -------------------------- Call for patent disclosure: - None ------------- Review of ARs: - Arpad update Table BIRD - Done - Walter send Time Domain GetWave Flow diagram for posting - Done - Walter send Backchannel proposal for posting - Done - Bob write a BIRD on correcting Table 1-3 in the spec. (Row 23). - In progress, on hold ------------- New Discussion: Arpad showed the Boolean BIRD draft: - Bob: The only question is if it is case sensitive - We have no Boolean type, this BIRD adds it - Arpad: It still supports string, and that can be used - Bob: Not tool supports this because we didn't define it - It could be a string "True" or "False" - Arpad: Boolean is already described in the IBIS spec - Walter: This is an over sight in section 10 - BNF stands for "Backus-Naur Form" - Bob: We have not defined the pipe character for OR Arpad showed the Format Corner and Range BIRD: - Arpad: Do we agree the GUI lets the user set typ/min/max corner? - No disagreement - Walter: I use typ/slow/fast to avoid the confusion over mapping - Arpad: How do we know whether "slow" maps to "min" or "max"? - Fangyi: The reverse problem already exists, mapping min/max to slow/fast - Todd: The reverse is the bigger problem - Walter: IBIS already has the typ/slow/fast concept - Bob: They map to the same thing, practically speaking - Todd: TTL gets slower when hotter - Semi vendors specify typ/slow/fast - I like that better - Bob: It means weak/slow, strong/fast - Arpad: Does Range have typ/min/max too? - Walter: Range gives a choice between 2 values - Todd: It is continuous - Ambrish: There is no defined increment? - Todd: Right - Bob: Only one value is sent to the DLL - My understanding is it has an increment - Walter: Every format defines allowed values - Todd: A small set of values would use List type - A larger set would use increment - Arpad: Page 147 has only slow and fast, no typ - These are selected by position - Bob: That is a mistake Arpad showed the AMI Function Return BIRD draft: - We must clarify what is returned when there is no data - Failure=0, success=1 - Walter: There should be a message mechanism to give the reason for failure - That may be beyond the scope of this meeting - Arpad: Some tools use success=0 Arpad showed the AMI_parameters_out BIRD: - We should have a 3.2.1.7 section for AMI_parameters_out - Input should be optional - Ambrish: It should be optional - Bob: Is Init required? - Yes, but no parameters are required - Arpad: Are we allowing empty strings? - Walter: Code assuming a string is coming would expect a pointer - Arpad: Do we allow NULL pointers? - Something has to be passed to Init - Walter: NULL pointer should not be allowed - Mike: Who frees the string? - Arpad: That is already addressed - Walter: Maybe we should allow NULL - Small number of tools, large number of models - It would be one less restriction for model makers - The DLL deallocates all memory on Close - Todd: Agree it's OK for simulators to ignore NULL pointers - Curtis: This is a pointer to pointer - The tool can write something back - Todd: Should we set the pointer to zero? - Walter: BIRD 128 says that - Arpad: So input strings are non-NULL but output pointers can be - Curtis: When we say optional it doesn't mean the function argument can be left out Arpad showed the Table BIRD draft: - Radek asked to clarify the requirement for at least one template row for Usage out - The size can be different between calls - The entire table must be passed on each call - Bob: Agree with this change - Arpad: Can we vote? - Ambrish: Would like one more reading - Arpad: We will vote next week Arpad showed the In/InOut BIRD - Arpad: The sentences have been reorganized - The only technical change is discussion of Usage Info - Has restriction on affecting simulation results - Walter: We agreed this would be 5.1+ - We have not defined what simulation is - Scott: It included post-processing, necessary in statistical mode - Walter: I will have something to discuss on this - Would like to table this until then - Arpad: We have a specific problem that this fixes - This is version 10 of this BIRD - Would like to finish it - Bob: We did not decide on this specific issue in Open Forum - Walter: We decided Friday this would go into 5.1+ - Bob: We didn't vote on that list - Walter: It's OK if we can also address the Jitter BIRD, etc. - Scott: Model_specific parameters have been used for post-processing because they were not well defined - We would also want to consider Jitter - That could be in 5.1 - Bob: We may have to cut off some BIRDs for logistical reasons - Scott: We can't go to 5.1 with things undefined - Arpad: We discussed taking two steps - Clarifications and reformatting would go into 5.1 - New features would go into 5.2 - One would follow the other quickly - Scott: This BIRD has to be deferred to 5.1+ then - It creates a loophole for post-processing - Arpad: I would like to get this BIRD out of our group - Walter: My presentation may make a difference ------------- Next meeting: 24 May 2011 12:00pm PT Next agenda: 1) Task list item discussions ------------- IBIS Interconnect SPICE Wish List: 1) Simulator directives