IEEE P1076.1 Working Group - Table-Driven Modeling Project - Telco Metting Minutes of July 5, 2010

Status: approved on August 9, 2010.

Attendees

• Ernst Christen
• Joachim Haase (chair, meeting convener)
• Arpad Muranyi
• Alain Vachoux (note taker)

Agenda

• Call to order
• Approval for agenda
• Review and approve minutes from June 7 meeting
• Review action items and any content provided from them
• Discussion of TDM Requirements document
• Next meeting
• AOB
• Adjourn

Call to order on July 5, 2010 at 8am PST (5pm CEST)

Joachim welcomes the participants and starts presenting the meeting slides.

Volunteer for minutes

Alain volunteers to take notes.

Approval for agenda

Agenda unanimously approved.

Minutes from last meeting

Minutes of June 5, 2010 meeting unanimously approved.

Current action items

Alain:

• Register Arpad to the Twiki site. - done

Arpad:

• Provide links to places in IBIS web site related to table-driven modeling. - done

Discussion (see Tables in IBIS and IBIS Specification version 5.0):

• Two kinds of tables: I-V tables (DC) and waveform tables (transient).
• Not fully populated grids are supported.
• Only linear interpolation is required.
• Tables have size limits (I-V: 100 points, waveform: 1000 points).
• Suggestion to have parameters for lookup functions such as pin, model, etc.
• Need to index waveforms with V_/R_/L_/C_fixture. R_fixture and V_fixture are the most important ones.
• Most tools support up to 4 waveforms in a model (pull-up, pull-down, turned-off, turned-on).
• No need to necessarily implement the full IBIS table processing.
• Need examples in IBIS 5.0.
• IBIS models include more information than table data.
• Arpad implemented a prototype IBIS to VHDL-AMS Perl translator.

Joachim:

• Provide information related to table-driven modeling from the Modelica environment. - done

Discussion (see Tables in Modelica):

• The so-called "temporary functions" must be replaced by other functions in the future.
• The so-called "CombiTableXX" are components that must be instantiated the model.
• More details on the Matlab 4 storage format are needed. However, it is more important to define how to interpret data. File format is a secondary issue.
• The "ContinuousDerivative" smoothness is related to linear and cubic splines.
• Periodic extrapolation is supported in Modelica and should be added to the requirement list.
• Support of step functions (discontinuities) not possible if data grid is not required to be ordered. Supporting the nearest/closest data point interpolation method could however help.

Ernst:

• Look at the requirements and consequences of supporting multiple dependent variables. - open
  
• Look at how Verilog-AMS control strings are actually working. - in progress

Discussion (mostly on table support in Verilog-AMS; see Cubic spline interpolation in Verilog-AMS):

• Ernst did an inventory of the support of control strings in different versions of Verilog-AMS and different tools. No result in well organized form yet.
• Quadratic spines may have oscillation problems.
• Multi-dimensional spline interpolation as supported in Verilog-AMS uses relatively simple algorithm.
• Data point ordering may be costly be it must be done once and should speed up the accesses and the interpolation.
• We must define the format of the conditioned (e.g. ordered) tables and mappings from standard file formats.

Alain:

• Update requirement document from the meeting discussion. - version 0.2 published on web site

Discussion:

• In the 1D case closest data point method (interpolation) and periodic extrapolation method (cmp. Modelica) should be supported.
• Independent variables of table functions shall be of real type (scalar or vector).
• The approach shall support dependent variables of real type. It may support complex dependent variables.
  
• It has to be clarified what are the consequences if real arrays are supported by the functions.
• The consequences of the requirement to ignore changes of data points after the first call (TDM-R11 in v0.2) have to be investigated. Thus, the requirement should be removed fro the current version of the requirements.
• It was recommended to reformulate requirements concerning the description of the data base and the formats that should be supported.

Next meeting

The next telco is planned on August 9, 2010 at 8:00am PDT or 5:00pm CEST.

AOB

None.

Adjourn

Joachim adjourns the meeting at 6:30 CEST.

Action items

Alain:

• Update requirement document from the meeting discussion.

Arpad:

• Provide examples of the use of tables with IBIS 5.0.
• More details on the prototype IBIS to VHDL-AMS Perl translator.

Ernst:

• Look at the requirements and consequences of supporting multiple dependent variables.
• Writeup on the use of Verilog-AMS control strings with different language definition versions and tools.

Joachim:

• Details on the Matlab 4 storage format.