From - Wed Aug 30 17:09:51 2000 Received: from mailgate2.Cadence.COM (mailgate2.Cadence.COM [158.140.2.31]) by zip.Cadence.COM (8.9.3/8.8.5) with SMTP id MAA03897 for ; Wed, 5 Apr 2000 12:24:35 -0400 (EDT) Received: from thalia.fm.intel.com (thalia.fm.intel.com [132.233.247.11]) by mailgate2.Cadence.COM (8.9.3/8.9.3) with ESMTP id JAA01448 for ; Wed, 5 Apr 2000 09:24:30 -0700 (PDT) Received: from SMTP (fmsmsxvs01-1.fm.intel.com [132.233.42.201]) by thalia.fm.intel.com (8.9.1a+p1/8.9.1/d: relay.m4,v 1.19 2000/01/29 00:15:43 dmccart Exp $) with SMTP id QAA16325; Wed, 5 Apr 2000 16:25:14 GMT Received: from fmsmsx27.FM.INTEL.COM ([132.233.48.27]) by 132.233.48.201 (Norton AntiVirus for Internet Email Gateways 1.0) ; Wed, 05 Apr 2000 16:24:29 0000 (GMT) Received: by fmsmsx27.fm.intel.com with Internet Mail Service (5.5.2448.0) id ; Wed, 5 Apr 2000 09:24:27 -0700 Message-ID: <4575832C8E71D111AC4100A0C96B512706458A06@fmsmsx36.fm.intel.com> From: "Muranyi, Arpad" To: "'Mike LaBonte'" , Al Davis , "Coleman, Dave" , "Peters, Stephen" , Scott McMorrow , Weston Beal , John Angulo , "Muranyi, Arpad" Cc: Bob Ross , "Mellitz, Richard" , "'dc.sessions@vlsi.com'" , "'lgreen@hyperlynx.com'" Subject: Minutes of IBIS Futures subcommittee meeting, March 30, 2000 Date: Wed, 5 Apr 2000 09:24:25 -0700 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2448.0) Content-Type: text/plain; charset="windows-1252" X-Received: By mailgate2.Cadence.COM as JAA01448 at Wed Apr 5 09:24:30 2000 All, Please let me know if I forgot to include anyone on the mailing list, and if there is a need to correct anything in the minutes. Thanks, Arpad ===================================================================== Minutes of IBIS Futures subcommittee meeting March 30, 2000 8:30 - 5:00 Intel (JF1201) Portland, OR Next meeting: Decided to continue with discussions on EMAIL and tentatively plan on meeting in about a month. Participants: ============= John Angulo Hyperlynx Weston Beal Mentor Dave Coleman Intel Al Davis Hyperlynx Mike Labonte Cadence Scott McMorrow SiQual Arpad Muranyi Intel Stephen Peters Intel 8:30 Arpad started the meeting with going through the responses to his "Thought provoker" message he posted on the subject of IBIS futures in December, 1999 on the SI-list reflector. Free discussions took place on some of the subjects as we went along. Some of the more noteworthy remarks were: - IBIS is still useful, and we still need a standard. - Currently IBIS consists of predefined structures where users supply data/parameters to this structure. Now go to more of a create a model approach. Right now "developer spends time on algorithms, not on data interpretation." - Public key encryption will be inevitable (and appropriate) because we are getting to that level of detail. Automatic SPICE to IBIS translation wont hide IP. One possible ways would be that each tool vendor supplies model maker with their own encryption program for making vendor specific encrypted model. Another option would be to have public/private key, administered by IBIS committee. - establish a subcommittee to deal with the subject of encryption - Discussion on XML as a format has some advantages (parsers and software available). - Need backwards compatibility? No, not necessarily. - Fundamental capabilities of description language: * reuse of instances * global variables w/ override * export of local variables * named block hierarchy * library reference (from external library) Stephen Peters wrote two list on the white board: Requirements of the new spec. - standardized description - protect IP - support for transmission line simulators - support circuit level simulators - document SI parameters - support pseudo static buffer configs (change I/V curve depending on DC loading) Things to fix or add to old spec. - receiver modeling - account for multi output switching effects - account for effects of pkg on output switching waveforms - support general/enhancable method for specing measurements at any node - support ability to model die interconnects (power delivery) - enable S-parameter equations, descriptions - support lossy transmission line stuff 10:30 Arpad presented his ideas for the new nodal syntax - add [Pad] keyword for listing all of the die pads - remove buffer mapping from [Pin] keyword - use a spice like element syntax without the "First Letter" element definition, describe only what happens between nodes with constants, variables, expressions, PWL tables, etc. 11:15 Al Davis's macro language - addresses prototype models - allows existing IBIS models to be translated to new syntax - backwards compatible - does not address pkg stuff or hierarchy - can we sell this? 12:00 Lunch =========== 12:45 Stephen Peters' Presentation - problem with 1-to-1 mapping between pin and pad 13:45 Free discussion Scott McMorrow noted that we need a general way to connect pins to internal models. - [External Pins] list of pins (could have signal name for documentation) - [Interconnect] how pins interconnect to pads/actual models 14:30 Mike Labonte brought up transmission matrix format problem - which is better, the more compact format HSPICE uses (L, C, Ro, Rs, Go, Gd) or a full matrix for multiple frequency points (Lf, Cf, Rf, Gf)? - concluded that the more compact format loses some accuracy because the R=Ro+Rs*SQRT(f) equation is not exact - however, field solvers which do not include loss tangents do not provide perfectly accurate results either, so matrixes at multiple frequencies are not perfect either 15:15 Discussed details of Al Davis' proposal - the need for a validation section, reference waveform came up - add a library calling mechanism - the need for a stimulus pattern definition came up - model_type confusion needs to be fixed Al defines a "model types" as his prototype circuit, but "model_type" also defines whether model is IN, OUT or I/O. - include timing constraints in model spec section? We all liked Al's proposal for the modeling section, but acknowledged that it didn't address the non-modeling related features (wrapper). 4:30 Summary - external pin / interconnect - nodal style syntax - subcircuits - Al's macro language - define interconnect and element syntax - timing (AC sheet information / constraint data) in IBIS? - make it case insensitive? AR: Al Davis, make some refinements on he proposal per our discussions Al Davis, write a proposal for wrapper syntax around the model Arpad Muranyi, write a proposal syntax for the header section 5:00 Meeting ended ==========================================================================