Dan,
Your imitation of the totalitarian Nikita
Khrushchev was a good joke and people will understand that you were having
fun. Like all good jokes, it was based on stretching
the truth. The 100GE ONLY camp wants to set the standard the
way it's always been done with a 10X leap and that's the only way they will
accept it. We're lucky that the IEEE has a strict rule of 75% approval
for the PAR, so that consensus needs to be built to move the standard
forward. I feel that minorities such as associates from Intel, Broadcom
and SUN should have a say in the future of the HSSG.
The 40GE camp has never claimed 100GE is invalid.
They claim that the HSSG - that is defining speeds beyond 10G - is invalid
without 40GE. This is the HSSG project, not the 100GE
project.
One characteristic of totalitarians is that they
deny obvious truths. One claim you made in
Geneva was that 40GE has not proven/shown any PMDs. Since the
40GE camp is proposing to use a subset of the 100GE PMDs, does the 40GE group
need to prove the capability of doing less than 100GE?
Can anyone argue that doing 4 lanes of 10G is not
technically feasible while doing 10 lanes of 10G is?
There are two standards-based transceivers that should be
out within a year that qualify as 40GE transceivers - the QSFP and the
X40. No 100GE transceivers are being standardized, so only
proprietary solutions are available at 100GE. Of course we could have
standard 10X10G solutions by 2010, but good luck creating a 4X25G
standard by 2010.
Others have claimed that the 40GE camp has
not met the 5 criteria of the PAR. Howard Frazier showed how
40GE met the 5 criteria in frazier_03_0507. Are there objections to this
presentation?
Another objection is that 40GE will cause confusion in
the market place. With distinct PMDs, the customer should be able to
distinguish a 40GE port from a 100GE port. The customer will not be able
to plug cables willy-nilly, but they shouldn't do that anyway on mission
critical systems. Interoperability concerns will not bring links down,
but users will not be able to bring links up that are not properly
cabled/configured.
The main suggestion of the 100GE ONLY camp is that 40GE
should create its own PAR. Besides doubling the work for the IEEE
standards approval process, the main concern is that the 40GE PAR would not be
approved for the reasons that the 100GE ONLY camp is currently stating.
This strategy of kicking 40GE out of the HSSG and then nipping it in the bud
in the PAR process has been described as the divide and conquer
strategy. Keeping Ethernet speeds above 10G that are very similar in
characteristics (but not PMDs) is the most straightforward process.
The ugly and distorted path is having two projects trying to proceed
at the same time when one project is a simple subset of the other
project.
Can 40GE be included in the HSSG PAR if it is a simple
subset of the 100GE PAR? This would require minimal standards effort as
shown in frazier_04_0507 and muller_01_0407.
If the 100GE ONLY camp has
a super-majority of 75%, then 40GE will be forced to move on to
its own PAR. While we still operate under the rules of the IEEE, the 25%
minority can not be thrown under the wheels of 100GE. The free markets
are better at determining the merit of a product than tens of people in the
HSSG.
Kind regards,
Scott Kipp
QSFP Chair
Office of the CTO
Brocade
Hi Chris,
Thanks for your concern about my frustration with the way
things were going in Geneva. After a week of touring Rome, Tuscany and the
Cinque-Terra National Park with my family, I feel totally rested and relaxed
and ready to get back to work. After a
relaxing week in Italy, I still think that my frustration with the debate
in Geneva was justified.
As I said in my presentation, it was clear that a
super-minority was willing to stall 100G progress in an effort to get an
unprepared and unjustified 40G project through the process. The debate after
my presentation, including last-minute additions to their objectives to
gain a few votes demonstrated that I was on target completely.
Fortunately, the record shows this as well.
In hindsight, I have only one regret with regard to the
presentation I gave on Thursday morning. The photograph I inserted as an
attempt to provide some levity into a serious discussion (which I am known to
do) might be misinterpreted by someone who was not in the meeting and
thus unaware that it was NOT a real part of the debate, but rather a staged
photo designed to make light of the meeting location which was taken
before the debate began. I should have left that out of my presentation so
that the record would not contain any ambiguity as to the
completely professional approach I took to the
debate.
Now, how do we move forward?
I think the path is relatively straight forward if we
choose to take a straight forward path. It can get ugly and distorted, only if
we choose to take an ugly and distorted path.
The record shows that 100G Ethernet has been justified
as a project for the IEEE 802.3. We have shown economic feasibility,
technical feasibility, distinct identity, compatibility, and broad market
potential. Up until Geneva, there were only a few that would argue against
this. I respect the difference of opinion held by those who have issues with
the technical or economic feasibility of 100G, but they were a very small
minority and their concerns do not merit delaying the
project.
In Geneva, when I raised the motion to clarify this
point, the most vocal argument made against 100G being
proven was "Without 40G in the PAR, these conclusions are no longer
valid". So all of the work to justify a 100G project that was done prior to
Geneva, and voted on with overwhelming concensus, suddenly lost its merit? I
don't think so. I think it was a transparent argument posed to rationalize the
minority demand to add 40G into the PAR. There was little argument (and
only by that small but persistant minority) against 100G being proven as a
stand-alone project by the work we have done.
The straightforward path is obvious. The HSSG should
forward a PAR for 100G Ethernet as it has been written, reviewed, and approved
by a super-majority of the HSSG prior to Geneva. The super-minority should
recognize that stalling a well developed PAR will continue to be perceived by
the majority of the HSSG, our customers, and by outsiders, as
counterproductive.
In addition, as I initially proposed and subsequently
demonstrated, I am open to studying 40G as a server interconnect solution.
Consistent with my first presentation, we should consider it as a
separate PAR and perhaps in a new study group focused on that market need.
Such a project would have to be shaped to ensure that was
economically feasible, distinct, and that it would not result in
market confusion or an unjustified amount of standardization work. I think
this is a reasonable set of criteria for advancing a project and to protect
our customers and the industry from yet another minority-driven compromise
that forces the industry and the market to make a decision we did not have
the discipline to make ourselves.
Dan
Dan,
While I
share some of your frustration with the lack of progress during this week’s
HSSG meeting, and some of the individual points in your presentation, I do not
support it. I am in disagreement with your Post-Debate Conclusions, and find
their tone as not conducive to good discussion of the best way to move forward
within the HSSG. I am aware of the passions generated during this week’s
debate and understand why you wrote your presentation, but wish that it had
not been sent out. I am confident that after your vacation travel in
Europe with your family during the next week,
you will return to your typical constructive approach to resolving difficult
issues. I am also optimistic that you will take back your Post-Debate
Conclusions, and return to your very reasonable Pre-Debate Conclusions. At
that point, you can list me as one of your supporters. I am also hopeful that
other supporters of your presentation will do the same.
Good discussion of how to move
forward is critically dependant on acknowledging that 1) 100GE Broad Market
Potential, and 2) 40GE Broad Market Potential have been established well above
the threshold for 802.3 Five Criteria. Continued debate of this will only lead
to delay in addressing the substantive issue of what is the best way to move
forward in developing 100GE and 40GE standards. A possible framework for this
discussion is outlined in “HSSG Next Steps Proposal” presentation that HSSG
participants authorized as a post-deadline meeting
submission.
http://www.ieee802.org/3/hssg/public/may07/cole_03_0507.pdf
An insight that has come out of
the HSSG discussion of 100GE and 40GE rates during the past several months is
that fundamental development cycles for new network equipment architectures
and new server architectures appear to be different. A new data switch
architecture development requires massive investment, which leads network
equipment developers to want large jumps in data rate, like factors of 10x, to
allow a return on that investment. An intermediate data rate causes an
increase in the overall development investment, and a shortening of the useful
life of network equipment which reduces the return on that
investment.
Economics of server development
appear to be different and more favorable to shorter development cycles, i.e.
more frequent architecture changes. This drives the need for more granular
jumps in data rate, like 4x. Ethernet has not done this historically, but that
may be because protocols other then Ethernet were used to bridge the gap.
Going forward, more granular Ethernet data rate steps may become the norm,
rather then just a one time anomaly. This difference in development economics
is also consistent with how often some end users replace servers versus
networking equipment. It suggests that moving forward, network data rates may
go from 10G to 100G to 1T, while server data rates go from 10G to 40G to 100G
to 400G to 1T. This difference in data rate needs is not necessarily bad for
either industry. More frequent server replacement may extend the useful life
of network equipment (good for network equipment developers), and large jumps
in network data rates assures availability of aggregation capacity to support
multiple server cycles (good for server equipment
developers.)
As we discuss how to best move
forward with developing 100GE and 40GE standards, any approach needs to have
the following two characteristics; 1) permits network equipment developers to
have a single 100GE architecture, i.e. does not force them into developing a
dual rate 100GE/40GE architecture, and 2) gives server equipment developers a
40GE server data rate. It may also need the recognition that some data rates
are optimized for server interconnect and not intended for
networking.
I look forward to a constructive
discussion on how best to move both standards
forward.
Chris
Subject: [HSSG]
Soliciting Support for my Presentation
Date: Thu, 31 May 2007
05:25:12 -0000
|
Hello,
Subsequent
to yesterday's debate, I feel it is necessary to address some of
the issues raised in a formally documented
manner.
I have
produced a presentation which I have requested authorization to present
at today's meeting. If you support the content of this presentation,
please reply back to me asap.
http://www.dovenetworking.com/Dove-02-May-HSSG.pdf
I am open to
feedback as well.
Best
Regards,
Dan
Dove
----
Principal
Engineer, LAN PHY Technology
Dove Networking Solutions - Serving
ProCurve
916-785-4187 (Roseville Office)
530-389-9315 (Home
Office)
530-906-3683 (Cell)
http://www.dovenetworking.com
|