Ambrish,
I agree that BIRD 190 only applies to Time Domain and not Statistical,
although Time Domain does use the output of Rx2 AMI_Init if any model has
GetWave_Exists is False.
I have other issues with:
Note: The Rx2 executable model file writer for the downstream channels
with Redrivers should keep in mind that the impulse response that is
presented to the Rx AMI_Init function does not include the effects of the
upstream equalization. Therefore, the Rx AMI_Init function will not be
able to perform accurate optimization in the absence of the upstream
channel characteristics and/or equalization effects. For this reason, the
parameters of the Rx AMI_Init function should always default to valid
values or have a mechanism to accept user-defined coefficients and allow
the user to turn off any automatic optimization routines to ensure
successful simulations.
First, the writer of a "terminal" Rx AMI model does not know if it is the
end of a Tx/Rx channel or if the board designer inserted a Redriver
because of the loss of the channel. Model writers should assume that the
inputs to the model are valid, but should always include the option to
turn off any automatic optimization routines, and set user defined
coefficients. This has always been true for traditional Tx/Rx channels
when the Tx is GetWave-Only.
The statement "the impulse response that is presented to the Rx AMI_Init
function does not include the effects of the upstream equalization. " is
true according to the IBIS 6.1 flow is correct, but this flow gives the
wrong results. As a matter of basic principles, we should not re-enforce a
flow that gives wrong results.
Why not include something like the following?
If an EDA tool does not include the effects of upstream equalization as
specified in the IBIS 6.1 Redriver Flow, the simulations will be
successful, but can give the wrong results.
Walter
Walter Katz
<mailto:wkatz@xxxxxxxxxx> wkatz@xxxxxxxxxx
Phone 303.449-2308
Mobile 303.335-6156
From: Ambrish Varma [mailto:ambrishv@xxxxxxxxxxx] ;
Sent: Friday, June 23, 2017 4:35 PM
To: wkatz@xxxxxxxxxx; IBIS-ATM <ibis-macro@xxxxxxxxxxxxx>
Subject: RE: Open Forum Presentation
IBIS_6.1_BIRD_190_BIRD_Fangyi_BIRD_166.pdf
Hi Walter,
I believe you are missing the point. The BIRD does not claim to solve
these issues. It merely warns the model maker 'Please do not expect EQ or
IR of the upstream channels/models'.
That's it.
Thanks,
Ambrish.
From: Walter Katz [mailto:wkatz@xxxxxxxxxx] ;
Sent: Friday, June 23, 2017 3:56 PM
To: Ambrish Varma <ambrishv@xxxxxxxxxxx <mailto:ambrishv@xxxxxxxxxxx> >;
IBIS-ATM <ibis-macro@xxxxxxxxxxxxx <mailto:ibis-macro@xxxxxxxxxxxxx> >
Subject: RE: Open Forum Presentation
IBIS_6.1_BIRD_190_BIRD_Fangyi_BIRD_166.pdf
EXTERNAL MAIL
Ambrish,
I agree that what you wrote is only in the Time Domain section, although
the time domain section includes the case when all models have
GetWave_Exists=False, and also allows to use the IR output of Rx2 as one
of the options instead of doing deconvolution.
Walter
Walter Katz
<mailto:wkatz@xxxxxxxxxx> wkatz@xxxxxxxxxx
Phone 303.449-2308
Mobile 303.335-6156
From: Ambrish Varma [mailto:ambrishv@xxxxxxxxxxx] ;
Sent: Friday, June 23, 2017 2:27 PM
To: wkatz@xxxxxxxxxx <mailto:wkatz@xxxxxxxxxx> ; IBIS-ATM
<ibis-macro@xxxxxxxxxxxxx <mailto:ibis-macro@xxxxxxxxxxxxx> >
Subject: RE: Open Forum Presentation
IBIS_6.1_BIRD_190_BIRD_Fangyi_BIRD_166.pdf
Thanks Walter,
As discussed in the Open Forum today, BIRD 190 has no connection to the
Init only Statistical flow in the specification and hence I recommend to
include the 'Note' as described in BIRD 190 in the time domain simulation
flow for a Repeater link.
Thanks,
Ambrish.
From: ibis-macro-bounce@xxxxxxxxxxxxx
<mailto:ibis-macro-bounce@xxxxxxxxxxxxx>
[mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Walter Katz
Sent: Friday, June 23, 2017 2:14 PM
To: IBIS-ATM <ibis-macro@xxxxxxxxxxxxx <mailto:ibis-macro@xxxxxxxxxxxxx> >
Subject: [ibis-macro] Open Forum Presentation
IBIS_6.1_BIRD_190_BIRD_Fangyi_BIRD_166.pdf
EXTERNAL MAIL
All,
IBIS_6.1_BIRD_190_BIRD_Fangyi_BIRD_166.pdf
Walter
Walter Katz
<mailto:wkatz@xxxxxxxxxx> wkatz@xxxxxxxxxx
Phone 303.449-2308
Mobile 303.335-6156
