[ibis-macro] Re: Backchannel Comments
- From: Ambrish Varma <ambrishv@xxxxxxxxxxx>
- To: "ibis-macro@xxxxxxxxxxxxx" <ibis-macro@xxxxxxxxxxxxx>
- Date: Tue, 3 Sep 2013 11:40:06 -0700
Hi Walter,
Thanks for your note.
Backchannel training is a standard feature in most, if not all the standards
going forward. We have seen it in PCIeGen3, 10GBase KR and others (40/100Gbe
etc) so far and don't see it disappearing in the coming specifications as well.
AMI models need to cope with this or we are not going to be keeping up with the
technology.
As for the need for it, I am personally working with at least a couple of IP
vendors besides the one on the BIRD. The IP division at Cadence that is selling
IP based on all of the above standards today wants Backchannel to be supported
in their AMI models per the requests of their customers. These customers are
often using Cadence IP mixed with IP from other vendors.
As far as the cost of implementation is concerned, we are trying to solve a
technical challenge in this forum and not discussing business issues and how to
make money, albeit that is our ultimate goal.
All of IBIS developments come with significant costs which is why we always
push for general solutions and not favor continuous keyword addition. That kind
of churn adds to the bulk of the cost of development and maintenance. The more
we churn the spec, the more it costs the industry.
Best Regards,
Ambrish.
From: ibis-macro-bounce@xxxxxxxxxxxxx [mailto:ibis-macro-bounce@xxxxxxxxxxxxx]
On Behalf Of Walter Katz
Sent: Saturday, August 31, 2013 9:32 AM
To: IBIS-ATM
Subject: [ibis-macro] Backchannel Comments
All,
We got into the weeds a little bit on how stimulus patters need to be defined,
and we should be doing that - stimulus patterns are important, and require
flexibility in functionality and ease of input and parsing.
I would like to take a step back, and identify the cost of EDA tools
implementing back channel AMI models and IC Vendors need for such capability in
IBIS-AMI. As far as the cost to EDA vendors we can all make that judgment for
ourselves, but I think we can all agree that this is not simple, and require
man months of development and support costs.
As far as need goes, I have seen zilch form IC Vendors and Users, other than
the one supporter of this BIRD. I personally do not think we should waste any
more of our valuable time on back channel, until we get additional support from
the IC Vendor and User community, and then only after preliminary .bci files
are supplied for each SerDes standard that supports back channel training.
Walter
Walter Katz
wkatz@xxxxxxxxxx<mailto:wkatz@xxxxxxxxxx>
Phone 303.449-2308
Mobile 303.335-6156
Other related posts:
