Some comments from this peanut gallery:
I like "IBIS_AMI_Extensions" for the name. I'm aiming for something
politically neutral <wink>.
While enumerated "extension" parameters is a way for the EDA tool to parse
for ones it does or does not understand, it also can be rather unwieldy if
the feature involves several parameters, e.g. PAM4. Would EDA vendors
generally even bother to read the table (unless required to do so -- ick!)
or simply sqawk a bit if the Reserved_Parameter is present at all?
Not every possible extension is expressed as a parameter. I can think of
two already in use in my models. For example, one is a time-alignment of
the impulse response out of AMI_Init to inform the EDA tool exactly where
the recovered clock actually lies relative to the impulse response (we use
it with our internal tools -- it's better/easier than
Rx_Clock_Recovery_Mean -- but no other commercial tool uses it). In this
case we could tie it to an enabling/identifying parameter and perhaps any
such feature can be trivially signalled with a parameter.
On Tue, Sep 15, 2015 at 6:10 AM, Walter Katz <wkatz@xxxxxxxxxx> wrote:
Arpad,
I believe the consensus of IBIS-ATM is:
“ a way to communicate to the user that the model includes non-standard
features and will only work in certain simulators, the end user will know
what to expect from the model and how to use the model successfully.”
I have three comments on your proposal.
1. I would prefer a more a positive parameter name such as
*AdvancedParams* or* NonStandardParams* instead of *VendorSpecificParams.*
2. The parameter should be of Type Table to be consistent with the
way we have done similar Reserved Parameters (e.g. Supporting_Files)
3. It gives the EDA tool the ability to either
a. Ignore it.
i. The
User needs to be informed enough to read the .ami file, or look at the IBIS
Check report to find out if this model is using Advanced or Non Standard
parameters.
b. Inform the user that the model is advanced and this tool does not
use the advanced features.
c. Inform the user that the model is advanced, and this tool does
use the advanced features as described in ... .
Walter
-----Original Message-----
From: ibis-macro-bounce@xxxxxxxxxxxxx [mailto:
ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Muranyi, Arpad
Sent: Monday, September 14, 2015 11:55 PM
To: ibis-macro@xxxxxxxxxxxxx
Subject: [ibis-macro] Out_InOut_Info_BIRD draft_11
Hello Everyone,
Here is draft_11 of the Out_InOut_Info_BIRD we have been discussing in
recent meetings. This is a somewhat crude draft and I am open to
suggestions on what to add, take away or change. Please take a few minutes
to read if before the ATM meeting so we could have a productive discussion
on it.
I will also ask Mike L. to post this on the ATM website so it would be
archived and be available to everyone in case there are problems with the
attachment on this email reflector.
Thanks,
Arpad
=====================================================================
