====================================================================== IBIS FUTURES/COOKBOOK SUBCOMMITTEE MEETING MINUTES Date: April 7, 2005 Previous Meeting Attendees: ---------- Cisco - Syed Huq Intel - Arpad Muranyi Mentor - John Angulo Samtec - Otto Bennig Teraspeed - Bob Ross ====================================================================== 1 PM - 3 PM US Pacific Time Telephone Bridge Passcode 916-356-2663 1 534-1542 Agenda: 1 - 1:05 PM Opens - Bob 1:10 - 1:30 The parser license - Bob - discussion of the cost - $1,000 for commercial usage. 1:30 - 1:40 ibischk4.1.1 release - Bob/Syed 1:40 - 1:50 Possible Summit in Asia (China) - Bob 1:50 - 2:00 Cookbook - Arpad 2:00 - 2:30 BIRD97 - Arpad 2:30 - 3:00 BIRD95.3 - Syed ------------------------------------------------------------------------- Upcoming Futures Meeting: Date: Thursday, April 21, 2005 Bridge: 2 Passcode: 474-4324 ====================================================================== Futures Minutes (abbreviated discussion) Bob Ross temporarily ran the meeting until Syed Huq arrived later. ICM Parser License Bob Ross discussed the issues of the ICM parser license. The source code has been turned over to the IBIS Committee. Currently it is available under two GNU Public Licenses. These exclude anticipated realistic commercial usage, so We need a separate license for companies that want to embed the code directly in products (outside of the existing make the code public requirements). A new license needs to be created for such usage, and a fee in normally required to make this transfer legal. The suggested fee of $1,000 seems to be acceptable. Bob gave these licensing options: - simple like ibischk4 - more formal througth GEIA (Michael would handle this, if possible Bob suggested holding off until Michael Mirmak returns, if possible. He needs to make the call. However, if a more urgent needs for a new license arises, then we will act on it. We may need a formal release of the code from the developer. IBISCHK4.1.1 Release The ibischk4 V.4.1.0 parser release failed to include the BUG82 fix. This defect is reported as BUG87. Several people were involved in implementing the simple fix (adding again the left out fix.) Bob Ross reported that Jon Powell has created a new release. Several executables have been created: - linux (with the release, checked by Syed Huq) - sun_5/solaras_58 created by Syed Huq - dos32 for Windows (John Angulo - sent during the meeting - hp11 to be done be Guy de Burgh The source code package needs some cleanup of extra files and the creation of a specific BUG87 test case for the regression suite. JOn needs to do this so that his master is the same as what is being sent out. Bob expects the work to be completed in a few days. The new executables will be uploaded. The new source code package will be sent to the companies owning parser licenses. Syed Huq will make a reflector announcment when we are doing this. Syed Huq joined and ran the remainder of the meeting. IBIS Summit in Asia (China) Bob Ross reported that he got encouaging feedback when he suggested that Huawei (PRC) host an IBIS Summit Meeting for Asia (China). This is in response to the success of the JEITA-IBIS meeting and the the general comment that we might want to have some IBIS Summits in ASIA in the future. Bob also discussed with Barry Sullivan of the International Engineering Consortium (IEC) possible plans of an Asian DesignCon. These are in the longer term plans (possibly as early as 2007). Bob stated that we would certainly be interested in joining/supporting DesignCon in Asia. Chances are that the first Designcon will be in some other Asian country such as Japan. However, Bob plans to start the discussion with Huawei (as a possible major sponsor) of an IBIS meeting but keep Barry Sullivan in the loop. There are many logistical details to consider, so this may or may not happen for several years. Bob also plans to bring in other US people who have visited China as we progress. Syed suggested trying to get Huawei more active in IBIS meetings. COOKBOOK Arpad Muranyi has a later version of the Cookbook (Ver 0.7b) and will send it to Syed Huq for distritution. Arpad wants more feedback. Syed will ask for feeback an put is as an agenda item for the next meeting. BIRD97 Arpad stated the major feedback on BIRD97 was from Katja Koller to add a MOSFET based equation requiring a threshold voltage. The Threshold voltage is regarded as confidential by some companies. Also Arpad was concerned that this approach might apply only for MOSFET technologies, and not all technologies. Arpad suggested that a table based option could cover other technologies. Bob Ross mentioned just keeping the simple scaling opion might be good enough for practical purposes since it is already implemented in one product. Bob also had questions/comments on the scaling factor. Based on the discussion and comments from John Angulo, the terms need to defined better, the equation needs to relate to IBIS rails for [Pullup Reference] and [Pulldown Reference], the scaling formula needs to be corrected per intent - a portion (plus or minus) table adjustment converted to a multiplier. The coefficients can be used for all I-V tables including clamp tables (which could have on die MOSFET resistors (that change with rail voltages). Bob commented that BIRD97 was a substantial rewrite and left out some paragraphs in the original writeup. Arpad plans to start again after moving the existing contents to the ANALYSIS or ANY OTHER BACKGROUND Section. Bob mentioned some other formatting details. Finally, Bob was concerned about the typ-min-max format and rules. All corner parameters in IBIS require typ and default to typ. Arpad was concerned over this, and will think about it. BIRD95.3 Syed Huq mentioned the key suggested changes are to remove the *_equiv network. John Angulo also mentioned adding information linking to the [Power Clamp Reference] and [Gnd Clamp Reference] rails. These changes are planned in the next release.