====================================================================== IBIS FUTURES/COOKBOOK TASK GROUP MEETING MINUTES Date: October 13, 2005 Attendees: ---------- Cadence Design Systems - Lance Wang Cisco - Syed Huq Intel - Michael Mirmak, Arpad Muranyi Mentor Graphics - John Angulo, Ian Dodd Micron - Randy Wolff Teraspeed - Bob Ross ====================================================================== Next Meeting: Thursday, October 20, 2005 9 AM - 11 AM US Pacific Time Telephone Bridge Passcode 916-356-2663 2 304-1331 Agenda: 9 - 9:10 AM Opens 9:10 - 9:35 AM Power Delivery Topics - Final Case Comparisons 9:35 - 10 AM IBIS Spec. Next Steps ======================================================================== FUTURES --------- Power Delivery Topics ---------------------- Michael summarized the latest work in attempting to match voltages at the receiver between different inductance network simulations. Bob is suspicious of the continuing inability to match nearly identical cases. Arpad suggests that RMAX may be the controlling issue. For his own buffer from his IBIS class, RMAX < 1. This controls the internal timestep of simulation and may improve matching. Michael asked, how small is small enough? Arpad suggested using a Thevenin voltage/resistance pair to eliminate transistor simulation effects. Michael will attempt these changes. Next Steps for IBIS ---------------------- Michael opened the discussion by asking what BIRD improvements should be included in which specific IBIS versions. Bob suggested that IBIS 4.2 and 5.0 could be analyzed and voted in parallel. In terms of combining BIRD95 and gate modulation, Syed suggested that both should go in the next revision of the spec, whichever version that is. Bob favors the "quick approach" of issuing IBIS 4.2 immediately, with minor fixes, for the purposes of standardization. He recognizes that IBIS 4.x has no formal linkage to ICM, but this may need to be ignored for 4.2. He added that closing the loop on ICM, gate modulation, BIRD95 all together plus approval plus parser work is a two-year process. To delay standardization is to risk having non-standard 4.1-like implementations proliferate. Michael summarized the current list of pending features and improvements: BIRD95 Gate Modulation ICM Links BIRD74 User-defined measurements John Angulo inquired about External Model calls under External Circuit. Michael responded that the team hasn't seen a proposal yet, but it could be done. For a "quick approach" IBIS 4.2, Bob suggests extra sentences could be used to exclude Fig. 12 stuff. He noted that IBIS4.2_WIP has been uploaded, but that BIRD95 has been kept separate for the moment. Several on the team suggested that linking ICM using multi-lingual keywords could be the cleanest approach. There was agreement also on the approach of a simple implementation first of BIRD95, then adding more complicated concepts later. An assumption of 1-1 connections, with no branching and no extra keywords would be most simple. Linking the ICM and IBIS parsers is "out" at this point. ARs ----- Michael Mirmak - put slides together showing options on IBIS 4.2 and 5.0 John Angulo - analyze BIRD100 proposals, after Mentor discussion Sam Chitwood - analyze BIRD100 proposals, after Sigrity discussion Lance Wang - analyze BIRD100 proposals, after Cadence discussion