All,
Per Michael's request Attached is a marked up copy that contains my review
comment and suggest changes.
It makes some technical corrections due to omission or misuse of
terminology, and also corrects some terminology.
It clarifies some verbiage that [Interconnect Model Set Selector] contains
names for [Interconnect Model Set]s, not [Interconnect Model]s.
It also has some suggested editorial changes to simplify the presentation or
to cut some areas to remove dangerous and unnecessary duplication.
Bob
From: ibis-interconn-bounce@xxxxxxxxxxxxx
[mailto:ibis-interconn-bounce@xxxxxxxxxxxxx] On Behalf Of Mike LaBonte
Sent: Thursday, January 19, 2017 12:14 PM
To: Bob Ross; IBIS-Interconnect
Subject: [ibis-interconn] Re: Interconnect BIRD Draft 47 for team review
I wonder if "numerically higher position" might be mistaken, considering the
highest position in the list is numerically lowest:
1. [Package]
2. [Pin]
3. [Package Model] and optional [Alternate Package Models]
4. [Interconnect Model Set Selector]
We could change the wording to clarify, or change the list to work the way
we think, which is order of consideration:
The order of precedence for package model data to be used by EDA tools in
simulation is defined below, in order of most preferred to least preferred.
For a selected [Component], the first package format from the list below
that is available in the IBIS or related file shall be considered by the EDA
tool to be more detailed and therefore preferred.
1. [Interconnect Model Set Selector]
2. [Package Model] and optional [Alternate Package Models]
3. [Pin]
4. [Package]
Mike
From: Bob Ross [mailto:bob@xxxxxxxxxxxxxxxxx] ;
Sent: Thursday, January 19, 2017 2:52 PM
To: 'Mike LaBonte'; 'IBIS-Interconnect'
Subject: RE: [ibis-interconn] Re: Interconnect BIRD Draft 47 for team review
Mike and Michael,
We can delete [Component] because the rules of what keywords scope others
are defined elsewhere or will be shown in Section 3.1 diagram. For the
purposes of order of precedence for a [Component]:
The order of precedence for package model data to be used by EDA tools in
simulation is defined below, in ascending order for a selected [Component].
If a package data format at a numerically higher position on the list is
available in an IBIS or related file, that data shall be considered by the
EDA tool to be more detailed and is therefore preferred.
1. [Package]
2. [Pin]
3. [Package Model] and optional [Alternate Package Models]
4. [Interconnect Model Set Selector]
We do not need this, but a more detailed version of the above is
1. [Package] L_pkg, C_pkg, R_pkg required default Typ column entries
(and optional Min and Max column entries)
2. [Pin] L_pin, C_pin, R_pin optional pin_name-specific column entries
3. [Package Model] and optional [Alternate Package Models] references
to [Define Package Model] formats
a. Uncoupled paths denoted by [Number Of Sections]
b. [Model Data] descriptions for uncoupled or coupled paths described
by [Inductance Matrix], [Capacitance Matrix] and [Resistance Matrix]
keywords
4. [Interconnect Model Set Selector] references to [Interconnect Model
Set] descriptions containing [Interconnect Model] connections described by
IBIS-ISS and/or Touchstone files
----
(An [Alternate Package Model] selection overrides a [Package Model] entry
(required for the [Alternate Package Models] keyword to be listed) but we do
not need to get into that detail.)
Also, IBIS-ISS includes the some other "more detailed" formats such as
pole-zero, W-line.
Bob
From: ibis-interconn-bounce@xxxxxxxxxxxxx
[mailto:ibis-interconn-bounce@xxxxxxxxxxxxx] On Behalf Of Mirmak, Michael
Sent: Thursday, January 19, 2017 8:37 AM
To: Mike LaBonte; IBIS-Interconnect
Subject: [ibis-interconn] Re: Interconnect BIRD Draft 47 for team review
Mike,
Good question. [Package Model] simply identifies by name which package
model data is to be used for a given component, but the actual definition of
the package model data to use is provided using the other two keywords
mentioned on line (3). Those other two keywords are at the same level of
the keyword hierarchy as [Component]. [Package] and [Pin], however, are
hierarchically under [Component] in their entirety. We should probably
clarify this section on these points.
In addition, we do need to explicitly state whether [Interconnect Model Set]
should be at the same level as [Component] or not.
- MM
From: ibis-interconn-bounce@xxxxxxxxxxxxx
[mailto:ibis-interconn-bounce@xxxxxxxxxxxxx] On Behalf Of Mike LaBonte
Sent: Thursday, January 19, 2017 7:40 AM
To: IBIS-Interconnect <ibis-interconn@xxxxxxxxxxxxx>
Subject: [ibis-interconn] Re: Interconnect BIRD Draft 47 for team review
A question about page 8:
The order of precedence for package model data to be used by EDA tools in
simulation is defined below, in ascending order. If a package data format
at a numerically higher position on the list is available in an IBIS or
related file, that data shall be considered by the EDA tool to be more
detailed and is therefore preferred.
1. [Component]/[Package]
2. [Component]/[Pin]
3. [Package Model] (including [Alternate Package Models] and [Define
Package Model])
4. [Interconnect Model Set Selector]
Why are only two keywords explicitly listed as falling under [Component],
when in fact all of them are?
Mike
From: ibis-interconn-bounce@xxxxxxxxxxxxx
[mailto:ibis-interconn-bounce@xxxxxxxxxxxxx] On Behalf Of Mirmak, Michael
Sent: Wednesday, January 18, 2017 4:30 PM
To: IBIS-Interconnect (ibis-interconn@xxxxxxxxxxxxx)
Subject: [ibis-interconn] Interconnect BIRD Draft 47 for team review
Please find enclosed draft 47 of the Interconnect BIRD proposal. This
includes all the changes suggested during today's meeting. In addition,
numerous minor corrections and changes for consistency have been made.
Finally, Table and Figure numbers and captions have been added where
missing, using the numbering present in the current 6.1 IBIS document, but
not taking any other approved BIRDs into account.
Please review the draft in preparation for next week's meeting, particularly
keeping in mind the status update/report we will be making to the IBIS
Summit at DesignCon. Thank you!
- MM
