Subject: [IBIS] Draft IBIS ICM 1.0 Swath comments [LONG]
From: Mirmak, Michael (michael.mirmak@intel.com)
Date: Wed Oct 02 2002 - 22:47:52 PDT
All,
After reviewing the [ICM Swath Parameters] section of the specification in
greater detail, I have a number of comments and questions.
1) The examples in this section suffer from the use of numbers to represent
both the columns of the swath and the columns of the full interconnect. I
would strongly suggest the use of letters ("A," "B," etc.) for the swath
columns, to reduce confusion. For example, in the description of "Centering
the Swath around the pins of interest," the following sentences appear: "The
Swath above is centered about column 4. The paths of interest including
those of columns 3 and 5 are also described by swath matrix data." The
references to "column 4" and "columns 3 and 5" are not clearly either to the
swath or the full interconnect.
In addition, the examples should explicitly show the mapping of the swath
columns onto the full interconnect, preferably in a line below the full
interconnect diagram (see below for an example).
A B C D E
*--*--*--*--*
|\/|\/|\/|\/|
|/\|/\|/\|/\|
*--*--*--*--* Swath
|\/|\/|\/|\/|
|/\|/\|/\|/\|
*--*--*--*--*
*--*--*--*--*--*--*--*--*
|\/|\/|\/|\/|\/|\/|\/|\/|
|/\|/\|/\|/\|/\|/\|/\|/\|
*--*--*--*--*--*--*--*--* Full Interconnect
|\/|\/|\/|\/|\/|\/|\/|\/|
|/\|/\|/\|/\|/\|/\|/\|/\|
*--*--*--*--*--*--*--*--*
1 2 3 4 5 6 7 8 9
A B C C C C C D E Mapped Swath Pins
In the comments below, I will use the swath lettering scheme just described.
2) There is an ambiguity inherent in the description of "Expansion to Full
Sized Interconnect" under the "Repeat the center pattern" step. The text is
reproduced below.
"Starting at the center of the swath, the coupling pattern of just
the center column (column 3
[in which matrix?] in this example) is copied for the interconnect
matrix. In this case, the
coupling pattern extends forward to the next column. The pattern is
copied by mapping the
swath pin numbers into the interconnect pin numbers and copying the
center column forward
referenced data."
As written, this text permits at least three potential mappings of the swath
onto the full interconnect. The position that the "center column" of the
swath occupies when placed within the full interconnect is not made clear.
The meaning of "extends forward" is also not explained, so that the pattern
from the assumed starting point is open to multiple interpretations, as
shown below:
Full Interconnect Columns: 1 2 3 4 5 6 7 8 9
Swath Columns (mapping one): A B C C C C C
Swath Columns (mapping two): A B A B C D E
Swath Columns (mapping three): A B C D E ? ?
The above examples assume that columns 1 and 2 of the full interconnect have
already been matched to swath columns A and B, as described in earlier
passages in the spec. Because the mapping isn't explicitly shown and the
phrase "extends forward" is undefined, the precise mapping of the swath onto
the middle of a full interconnect is not clear.
3) Similarly, when the text continues on to show the right-side mapping, the
example remains ambiguous at best. At least three different final
interpretations of the mapping are possible:
Full Interconnect Columns: 1 2 3 4 5 6 7 8 9
Swath Columns (mapping one): A B C C C C C D E
Swath Columns (mapping two): A B A B C D E D E
Swath Columns (mapping three): A B C D E ? ? D E
4) The final paragraph under this section, beginning with the phrase, "A
similar process could be created for expanding..." is too vague on details.
Leaving the interpretation of a swath with an even number of columns open to
each EDA tool vendor makes several possibly inaccurate and conflicting sets
of results possible for the same set of interconnect data. This risks
making data exchange possible at the expense of data accuracy, which would
destroy the utility of the ICM specification. At the very least, a
subparameter should be added so that the model author can select, for
instance, a left- or right-side "bias" when expanding a swath.
5) For clarity, the example shown under [ICM Swath Pin Numbers] should
explicitly show the row and column indicies above and to the left of the pin
field pictured (see below):
C1 C2 C3
R1 A B C
R2 D E F
Similarly, the phrase "physically arranged" should be inserted after the
word "pins" in the text just above this example ("Consider a 2 row by 3
column Swatch matrix that was extracted for the pins [physically arranged]
as shown below.").
6) This section was constructed with only connectors in mind, as their
regularity can be assumed. Would it be considered appropriate, however, to
apply the swath approach to ribbon cables or similar non-connector
interconnects? This would be a highly useful application. However, if this
is to be allowed under ICM, the language of the Swath section should change
so that "pin," "connector," and similar words are not used exclusively of
other terms descriptive of general interconnects.
7) The paragraph beginning, "None of the values of these numbers may exceed
the number of rows..." refers to "numbers" without a context. To what
"numbers" does the sentence refer?
8) The grammar and expression used in this section need significant
improvement before I would be satisfied sending it to a wider public
audience. Besides the issues pointed out above, there are numerous
instances of incorrect number agreement, poor hyphenation and use of
undefined terms (for example, "The edges of the matrix represent the same
edge effects that exists [sic] in the full size [sic] connector"). In
addition, many of the examples and much of the text would be more
appropriately placed in a "Data Derivation" section, similar to what is
present in IBIS 3.2, rather than in the descriptive text for the keywords
and subparameters.
- Michael Mirmak
Intel Corp.
michael.mirmak@intel.com
-----------------------------------------------------------------
|For help or to subscribe/unsubscribe, email majordomo@eda.org
|with the appropriate command message(s) in the body:
|
| help
| subscribe ibis <optional e-mail address, if different>
| subscribe ibis-users <optional e-mail address, if different>
| unsubscribe ibis <optional e-mail address, if different>
| unsubscribe ibis-users <optional e-mail address, if different>
|
|or email a request to ibis-request@eda.org.
|
|IBIS reflector archives exist under:
|
| http://www.eda.org/pub/ibis/email_archive/ Recent
| http://www.eda.org/pub/ibis/users_archive/ Recent
| http://www.eda.org/pub/ibis/email/ E-mail since 1993
This archive was generated by hypermail 2b28 : Wed Oct 02 2002 - 22:58:09 PDT