Minutes, IBIS Quality Task Group

30 April 2019

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

ROLL CALL

Intel Corp.                         Michael Mirmak
Mentor, a Siemens Business        * Arpad Muranyi
Signal Integrity Software:        * Mike LaBonte
SPISim                              Wei-hsing Huang
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 April 23 meeting were reviewed.  Bob Ross moved to approve the minutes
with correction as noted, Mike LaBonte seconded.  Without objection the minutes were
approved.


ARs:
- AR: Bob Ross to ask developer for an IBIS-ISS parser estimate.
  - Not done yet.


Parser impact of BIRD199:
Mike LaBonte showed BIRD199.  Arpad Muranyi said an example for Rx_Receiver_Sensitivity
showed a negative value in IBIS 6.1, but that was fixed in IBIS 7.0.  Mike said
the IBISCHK source code had no value checking at all for Rx_Receiver_Sensitivity.
Bob Ross said a simple test would be to IBISCHK a bad file.  He noted the BIRD is
not approved yet. A parser change would be effective only for [IBIS Ver] 7.1 if the
BIRD passes.  There could be a warning for prior IBIS versions, however.  Arpad agreed
that a warning could be added to IBISCHK7, applicable to all IBIS versions with AMI.
Mike asked if any negative values were found in currently distributed IBIS-AMI models,
noting that it is problematic when currently passing models start to fail IBISCHK.
No one knew of any.  Mike asked if the BIRD should have a note about our decision to
have a warning.  Arpad said BIRDs usually do not specify IBISCHK changes.  Bob said
there should be a caution for pre-7.1 negative values, not a warning.  Mike suggested
writing an IBISCHK bug report, and having the Open Forum weigh in on whether it should
be a warning or caution.

AR: Mike LaBonte to write IBISCHK bug report to warn for negative Rx_Receiver_Sensitivity

IBISCHK700 parser development:
No report from developer.


New IBISCHK bugs:
Bob Ross said we needed to decide the disposition of existing BUG203.  The concern
was that assuming the existence of a pullup reference terminal might not be valid
for some model types.  Mike LaBonte said that even without a Pullup I/V table to
call for the terminal to be present, the C_comp_pullup constitutes a call for that
terminal to be present.  The voltage for the terminal is known, so nothing blocks
the addition of the terminal..  Bob noted that Walter Katz had suggested setting bug
status to "WILL NOT FIX" because it is not a problem.  Arpad Muranyi noted that IBIS
does not specify requirements for this, that an input can have no clamp I/V tables at
all, and that C_comp_pullup is associated with the terminal for [Pullup Reference].
But he agreed that the simulator could establish 4 supply terminals no matter which
of [Voltage Range] or [* Reference] is/are present.  Bob said we already give a
warning for an input with a pullup or Pulldown table.  Lance agreed.

Arpad believed that for I/O Model_types the parser did not check for presence of Pullup
or Pulldown.  Bob felt it did.  Arpad suggested changing the IBIS specification to
require I/V tables.  Bob disagreed, noting that some people create zero current tables
in some cases, and this might force them to do so.  Lance suggested adding a note or
info message to the parser for this, not a warning or caution.  Bob said we only have
note, not info. He felt it should be a caution or warning.  Mike suggested that cautions
should have been enabled by default, but he agreed that no change should be made now.
He also suggested changing the bug and trying again to classify it.  Bob said we had
three options: do nothing, print a note, or print a caution.  Lance agreed.

IBIS-ISS parser discussion:
Bob Ross said he would be asking the developer for a quote on a simple parser, noting
that EDA tools already have the complex parsers they need.  Arpad Muranyi agreed
that tools already have suitable parsers.  Mike LaBonte said our own parser would be
primarily useful for validating that no non-IBIS-ISS is present, but probably not as an
actual parser front end for a tool.  He suggested that there would be not much reason
for anyone to buy the source code.  Arpad said the source code might be valuable for
helping to understand the rules accurately.  Bob said  the parser could be integrated
into IBISCHK.  Mike agreed with that idea, noting that a -iss option would provide for
checking standalone .iss files.  He suggested adding ISS parsing to IBISCHK and raising
the source code price to $3000 to help fund the additional development.  Arpad said
IBISCHK should use the given file extensions to direct the type of checking performed.
Bob said ISS files have no required file extension.  Arpad said he primarily wanted to
insure that standalone .iss files could be checked, as well as .iss files referenced
when checking a .ibs file.


Bob Ross moved to adjourn. Arpad Muranyi seconded. Without objection the meeting ended.

Meeting ended: 12:07 ET

Next meeting May 7, 2019

