Minutes, IBIS Quality Task Group

23 February 2021

11:00-12:00 EST (08:00-09:00 PST)

ROLL CALL

Intel Technology                    Michael Mirmak
Micron Technology                 * Randy Wolff
Signal Integrity Software:        * Mike LaBonte
Teraspeed Labs:                   * Bob Ross
Zuken USA:                          Lance Wang

Everyone in attendance marked by *

NOTE: "AR" = Action Required.

-----------------------MINUTES ---------------------------
Mike LaBonte conducted the meeting.

Call for IBIS related patent disclosures:

- None


Call for opens:

- None


Review of previous meeting minutes:
Minutes from the February 16, 2021 meeting were reviewed. Randy Wolff moved to accept
the minutes.  Bob Ross seconded.  Without object the minutes were accepted.


ARs:
- Mike LaBonte to produce TSCK2.0.1_alpha3 executables with bugs 3 and 4 fixed
  - Done
- Bob Ross to write IBISCHK bug for Touchstone checking
  - No progress


NEW ITEMS:

New parser bug reports:
None.


TSCHK 2.0.1 development:
Mike LaBonte demonstrated changes he had made to tschk2. The changes were:
- Change version to 2.0.1
- Correct URL in documentation from eda.org to ibis.org.
- BUG4: Allow -canonical-v1 for V2 file if all [Reference] values the same within 1e-6.
  - R is set to the first [Reference] value in that case.
- BUG3: Add -precision N,N option to set precision for printing frequency and data values.
  - Default remains at 15 (column width 20), matching tschk2.0.0 behavior.
  - Column width is precision + 5 to accommodate non-precision characters (eg.: _.____e+06).
  - Setting any precision to 0 uses maximum precision seen in observed input data.
  - Error for -precision values < 0
  - Add observed maximum frequency and data precision to -describe report (to see what 0,0 will do).
- -describe change above causes regression test failures, will require golden data update:

Mike said the column widths might need to be 6 more than precisions, not 5, to allow for
leading +/- signs.  Bob Ross questioned the alignment of numbers, which seemed variable
from row to row.  Mike said the numbers were left justified, including leading minus
signs, so digits would not always start in the same column.  He demonstrated the new
-precision option, setting fixed precisions of 10, 5, and 1.  Randy Wolff questioned
fixed vs scientific format choices for the 1 digit precision setting.  It was using
scientific format for numbers like 10.  Mike said he could test an option that would
force fixed format, but he felt that the output was correct for 1 digit precision.
The number 10 in fixed format had 2 digits of precision, but 1e+01 was considered only
1 digit.

Bob noted that tschk2 -describe output had more precision than requested in the
network parameters.  Mike said those data were in real/imaginary format, and -describe
would always show full precision.  The -precision option did not apply to -describe.
Bob said the observed precision from the input was stated just above the numbers, and
that was confusing.  Mike felt the wording might be improved.  He suggested removing
the printing of calculated precision from -describe and instead having it printed to
stderr before printing canonical output to stdout, when precision 0 was used.  He said
that should not affect existing EDA tool uses of tschk2, because the -precision option
did not yet exist.

Bob felt that giving 0 as the -precision value to signify automatic determination
of output precision might not be conventional.  Mike felt it was fairly common for
0 to signify some default when 0 is not a legal value.  Bob asked if precision was
calculated using all data, not just the first frequency.  Mike said all data were used.
Bob said that would have to be documented in the help output.

Mike showed short and long help outputs. He asked if only long should be used.
We decided to make no change.  Bob asked if it would be clear that "<opts>" in the
help refereed to the description of "Canonical output options" below.  Mike said he
would change "<opts>" in the help to "[options]"

AR Mike LaBonte to change tschk201 precision handling as discussed

Bob noted the documentation still showed version 2.0.0.

AR Mike LaBonte to update version number in tschk201 source documentation to 2.0.1

Mike showed tschk BUG1, requesting clarification of an error message.  Randy said the
BUG1 Touchstone testcase file had multiple R values on the # option line, not standard.
Mike said tschk2 was written to flag any duplicate setting as an error.  We agreed to
use the suggested error message from the bug report.

AR: Mike LaBonte to fix tschk2 BUG1 and BUG2

IBISCHK development:

IBISCHK 7.1.0 contract:
Bob Ross reported no update on the contract draft.


IBISCHK7 source code feedback:
Mike LaBonte reported Cadence had replied expressing no requests for change.  Bob Ross
reported that Mentor also expressed no request for change.  There was a chance Siemens
might impose requirements at some point.


IBISCHK security fixes:
No discussion.


Interconnect model file checking
No discussion.


Tabled topics (no discussion without motion):
  - BIRD181.2


Randy Wolff moved to adjourn. Bob Ross seconded. Without objection the meeting ended.

Meeting ended: 12:08 ET

Next meeting March 2, 2021
