Bonjourno,
First, I am pleased to report that there is nothing that I saw that
invalidates our new interconnect BIRD.
There were a number of presentations in both SPI2016 and the IBIS Summit
on different approaches to IBIS Analog Buffer modeling. Some were amenable
to the IBIS table data driven methodology, and some were not. I could not
tell if the methodologies are "research" or realistic. It would be very
hard to convince me that these methodologies could be implemented in any
or our EDA tools without great effort. I think the only way we could
practically enable these methodologies within IBIS is to enable a DLL
interface within [External Model].
I discussed the status of Touchstone 2 with several people. The bottom
line is that there are so many tools (both commercial and internal) that
cannot or will not be modified to be able to read (or write) Touchstone 2
that Touchstone 2 will never be accepted. The only feature that is needed
is impedance per port. It is my strong recommendation that Touchstone 2 be
withdrawn, and a much simple Touchstone II standard be developed which
simply adds a new optional record that adds a simple new record that
allows for impedance per port. Touchstone II would continue the standard
file extension s<#ports>p. What is also needed is an optional "side" file
(specifically for interconnect Touchstone files) that allows for the
naming of ports, near end far end information, and differential port
association.
The organizers of SPI2016 did an extraordinary good job. The meetings
where well organized (about 110 people attended SPI2016). There were about
35 attendees to the IBIS Summit. The social highlight of the conference
was a tour and white glove service dinner at the Royal Palace in Turin.
Walter
Walter Katz
<mailto:wkatz@xxxxxxxxxx> wkatz@xxxxxxxxxx
Phone 303.449-2308
Mobile 303.335-6156
