====================================================================== IBIS INTERCONNECT TASK GROUP MEETING MINUTES AND AGENDA http://www.eda.org/ibis/interconnect_wip/ Mailing list: ibis-interconn@freelists.org Archives at http://www.freelists.org/archive/ibis-interconn/ ====================================================================== Next Meeting Wednesday, July 13, 2011 9 AM US Pacific Time Telephone Bridge Passcode 916-356-2663 92 443-3339 (for international and alternate US numbers, contact Michael Mirmak) LiveMeeting: Agenda: - Attendees - Call for patents - Review of July 6 Meeting Minutes - Agenda & Opens - Proposal to release Touchstone 2.1 with only filed TSIRDs - MCP Plans - Agenda for Next Meeting - Approve Next Meeting Dates: * July 20 * July 27 (?; MM may not be available) ====================================================================== Minutes from July 6, 2011 Attendees: ---------- (* denotes present) Agilent - Radek Biernacki* Ansoft - Denis Soldo Cadence Design Systems - Terry Jernberg, Brad Griffin, Dennis Nagle, Taranjit Kukal Cisco Systems - Mike LaBonte IBM - Greg Edlund ICT-Lanto - Steven Wong Intel - Michael Mirmak* Mentor Graphics Corp. - John Angulo*, Arpad Muranyi* Micron Technology - Justin Butterfield*, Randy Wolff* Sigrity - Brad Brim SiSoft - Walter Katz* Teraspeed Consulting Group - Bob Ross* Xilinx - Ray Anderson* ======================================================================== No patents were announced. During Opens, Bob Ross mentioned two additional proposals related to IBIS-ISS: * (from Arpad Muranyi et al) adding support for .if/.elseif/.endif * (from Walter Katz) the suffix "A" is "atto" except when "AMP" is used, forcing units to "amperes" The team reviewed IBIS-ISS review draft 1.4. A single issue was raised by Bob Ross: on the markup copy p. 54 and p. 58 show two different font sizes on the H(s) equation. Other issues discussed: - Arpad Muranyi and Radek Biernacki noted that the "MEG/Mili/M" suffices will produce the same atto/amp issue as discussed above. Walter Katz suggested that the Open Forum may discuss this. Bob Ross noted that changes to this portion breaks existing SPICE tools. Radek added that whether "amps" or "amperes" will cause different behaviors is unclear. - Michael Mirmak noted that the source material describing if/elseif/endif have been checked and would include only two pages of changes. Arpad replied that he could accept waiting for an IBIS-ISS 1.1 for these features. Radek noted that this was originally deferred and that is should be discussed in the Open Forum. - Michael raised the problem of version control. No version control exists in any SPICE implementation currently known. This may be needed if features are added or changed in updates to IBIS-ISS. Arpad suggested that at least one SPICE tool uses the format yyyy.mm (year, month) for releases. Bob suggested that the document include version, but wondered if every IC vendor needed version control. Michael suggested ".version=xx" as a useful alternative that may or may not cause existing SPICE tools to crash. Bob replied that this needs checking, to ensure that ".version" results in a warning only. Also, he asked whether this option would be placed at the netlist level or at the subcircuit level. Arpad and Radek moved to release IBIS-ISS 1.4 to IBIS Open Forum. Bob amended the motion to cover the H(s) font change requested. The motion was approved with no objections. Michael noted that the updated document will be review draft 1.5. He will release the document to the Open Forum through the appropriate reflectors. The team discussed Touchstone 2.1. Michael proposed that only the submitted TSIRDs be included in a released Touchstone 2.1. Bob Ross agreed, adding that the binary features should be added as a separate section, while the sparse matrix section could probably be added in-line with existing text. Document issues include example and equation renumbering. Radek asked whether the renumbering should be executed through formal TSIRDs. Michael asked whether Touchstone 2.1 should include port-to-node mapping plus links to IBIS and/or IBIS-ISS. Should this wait until an Touchstone 3.0? Arpad replied that he needed to talk offline with several folks regarding if/elseif support in IBIS-ISS and how it would interact with Touchstone. Bob noted that sparse matrices are great reductions of content as-is and so could be used as a method for mapping. Walter Katz suggested simply closing the document and voting it out to the IBIS Open Forum. He added that the document still includes two deficiencies: - no method for documenting nodes-to-ports (see comments at DesignCon from Eric Bogatin) - no transfer-function-based descriptions (only frequency-domain-based descriptions are supported) Bob Ross moved to defer the decision until next week. No objections were raised. The meeting was adjourned.