IBIS Macromodel Task Group Meeting date: 19 April 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 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 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: - Arpad can not make it next week. - Walter would like to give two BIRD presentations - We need to finalize the Out-InOut BIRD - Would prefer to be present for that - Ken: There is supposed to be a diagram for the jitter BIRD - Can someone conduct the meeting? - We will decide later -------------------------- Call for patent disclosure: - None ------------- Review of ARs: - Arpad submit Crosstalk Clarification BIRD to Open Forum - Done - Ambrish start a BIRD on task list row 25 - In progress, on hold - Bob write a BIRD on correcting Table 1-3 in the spec. (Row 23). - In progress, on hold ------------- New Discussion: Arpad showed the Out-InOut BIRD draft: - Ken: Is this on the website? - Arpad: No, Scott made some comments that leave it unsettled. - A few phrases are added in the 2nd paragraph of page two. - These put additional restrictions on how simulators may use Model_Specific parameters. - Arpad: Walter disagreed and asked to have his name removed form the BIRD. - Should we vote on this? - Ken: We should delay a vote until next week - Bob: Any tool-specific usage is not compliant with the AMI spec - John: If there is a "my model is super fast" parameter it should be allowed to change how the model works. - Mike and Ambrish agreed. - Radek: That directive would be for the model, not the tool. - Arpad: Only Input parameters can change model behavior - Mike: The first sentence should have "Usage Out or InOut". AR: Arpad update Out-InOut BIRD and send to Mike for posting Arpad showed a document from Bob: - Mike: There has been no request to post this so far. - Bob: - Concerns with Walter's proposal: - Added rules when we are just trying to describe Table syntax - Rows must be delimited by parentheses - If there is a single Type it applies to all columns - If multiple Types the number of columns must match the number of Types - Mike: Some languages allow the last Type to apply to any additional columns - Radek: Type could be defined at the Table level with new meanings - Arpad: We should not have the text dictate what the type is as Walter proposed - Bob: It is awkward to have Type as a leaf inside a Table - Label is different, not the the same level as Type - Type is a require parameter - Radek: It is OK now but the number of Types dictates the number of columns - Arpad: The wording for Label is good, could be used for Type - Bob: Type could be moved down to the same level as Label - Labels are informational, never sent to the DLL - Radek: The number of columns must be counted by parentheses - This determines the number of rows - Bob: The DLL may return a different number of rows - Arpad: For Out Tables the .ami format dictates what the DLL will return - This is important - Bob: The .ami gives the number of columns - Arpad showed the last example - Bob: The Labels documents the number of columns - Arpad: That template is only "suggested", not required - Bob: The implication then is that at least one row is required - Arpad: For Out parameters do we require the .ami to contain any data? - It could contain just ((Usage Out) (Type Integer)) - Bob: The spec does not allow that - Radek: There is no rule saying you may not specify in the Out - Arpad: What if the .ami conflicts with what the DLL produces? - Maybe the default makes no sense - Ambrish: The model dictates what the table actually contains - Radek: The .ami just gives a placeholder - Arpad: This will have to be continued in a new BIRD draft - Radek: The number of rows might be decided on the fly - The number of columns and rows is fixed, so the value is the Table - There is no parameter giving the number of rows - Arpad: The suggestion is to require 1 row where Usage is Out - Bob: We could generalize Tables if we want multidimensional tables - This was broken form day one - Arpad: We could have a number of rows and not require 1 sample row - Bob: Sometimes memory must be allocated in advance - Radek: Number of rows is a good solution - Arpad: This should be continued next week ------------- Next meeting: 26 Apr 2011 12:00pm PT Next agenda: 1) Task list item discussions ------------- IBIS Interconnect SPICE Wish List: 1) Simulator directives