Hi Paul,
Good question. It is not yet in standards, but rather it is an example of an entry in the informative table that Matt proposed we add to future standards.
Specifically, the 100GE-LR4 link budget is 8.5dB, of which 6.3dB is loss budget which can be used in a variety of useful ways like going over SMF or through passive loss elements like path panels or cross-connects. There is also a 2.2dB allocation for penalties. That is one use scenario. The example I gave shows another use scenario. If an end user only needs a 2km reach, they have available in round numbers,~7dB of loss budget. Obviously if we were writing these numbers into the standard, we would be more precise.
Keep in mind that link budget itself is informative. The normative specs are TDP and SRS. So Matt’s proposal to add an informative link budget table is entire consistent with how we handle the single link budget entry today. Chris From: Kolesar, Paul [mailto:PKOLESAR@xxxxxxxxxxxxx]
Sent: Monday, November 21, 2011 5:29 AM
To: STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_100GNGOPTX] Emerging new reach space Chris, Can you cite the location of the 2km / 7dB information in the standard? I only find information on 30 km vs 40 km engineered link for –ER4.
Paul
From: Chris Cole [mailto:chris.cole@xxxxxxxxxxx]
Sent: Sunday, November 20, 2011 3:31 PM
To: STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_100GNGOPTX] Emerging new reach space Matt, Adding an informative table in future standards on reach vs. loss budget trade-off is an excellent idea. Many end users do this implicitly already, so they will benefit by having more exact numbers from those that write the standards. For 100GE-LR4, this type of information is already informally provided, for example pointing out that for shorter reaches like ~2km, a ~7dB loss budget is available. Chris From: "Matt Traverso (mattrave)" <mattrave@xxxxxxxxx>
Date: November 20, 2011 12:36:21 PM EST
To: John D'Ambrosia <jdambrosia@xxxxxxxxxxxxxxx>, "STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx" <STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx>
Subject: RE: [802.3_100GNGOPTX] Emerging new reach space
Colleagues,
John wrote: "Also one lesson that we should learn from recent industry debates -forget the reach- what is the desired budget?"
As Chris Cole & others have pointed out, the reach is dependent upon the transmission penalties as well as the link budget -- see discussion on wavelength pitch. It might be a useful approach to consider trading off a scenario where the amount of additional insertion loss (connectors, splices etc) versus distance.
For example, we could set an objective of 1km with 2dB of insertion loss (no significance -- I just pulled it out of the ether). Then we could provide an informative table which could highlight at 500m reach the allowable insertion loss is 2dB + X. Alternately, we could provide a scenario which shows that if the insertion loss is only 1dB then the reach is 1km + Y.
On the HSSG reference, I believe it was Ted Seely coming to the microphone to object to the presentation Chris, Eddie & I put together (http://www.ieee802.org/3/hssg/public/nov07/cole_01_1107.pdf) -- he was concerned that there were many scenarios where there were higher connectors/insertion loss such that he preferred 10km to accommodate the additional insertion loss.
cheers
--matt
-----Original Message-----
From: John D'Ambrosia [mailto:jdambrosia@xxxxxxxxxxxxxxx]
Sent: Saturday, November 19, 2011 12:27 PM
To: STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_100GNGOPTX] Emerging new reach space
Paul
I notice you used a city block but you did leave out the fact that it was a multifloor scenario.
However as I pointed out I remember this from hssg days which is 5 years ago. I agree with the call for new data and am not sure what data that has been presented to the IEEE that details the industry 2km need. How much of that is want and not need is unclear to me.
Also one lesson that we should learn from recent industry debates -forget the reach- what is the desired budget?
John
Sent from my iPhone
On Nov 19, 2011, at 2:16 PM, "Kolesar, Paul" <PKOLESAR@xxxxxxxxxxxxx> wrote: Scott, I agree with the thrust of your assessment.
A minor correction. I did include the entire link distribution in my analysis. The only truncated input in my analysis is the patch cord distribution above 70 ft. The percentage of channels within (or above) any length is the result of convolution of the length distributions of constituent links and cords, which carries the probabilities with it. For example, if 10% of links lie above 100 m, the probability of two concatenated links being longer than 200 m is 0.10 x 0.10 = 1%. This is over simplified because it does not include the equipment cords and the patch cord that create full channels.
I note that your cut-off at 400,000 square feet allows for a square footprint that is 632 feet on a side, substantially bigger than a square block which, at one tenth of a mile on a side, is 279,000 square feet.
Paul
-----Original Message-----
From: Scott Kipp [mailto:skipp@xxxxxxxxxxx]
Sent: Saturday, November 19, 2011 11:59 AM
To: STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_100GNGOPTX] Emerging new reach space
Jeff,
I agree with your point about distributions.
No, it was not arbitrary that I stopped at 400,000 sq ft. I stopped at 400,000 sq ft because I wanted to cut off the tail of the distribution of the size of the data centers. I propose that there are only a few % (probably les than 1%) of data centers that are over 400,000 sq ft. The larger ones are the mega data center or massive data centers (MDC in either case) that Google and others are deploying.
Google is a very interesting case study since they deliver over 6% of all Internet traffic. I just got a response from Bikash and he says they do have 2 km links and he is getting frustrated with the IEEE second guessing this reality. What I think we need to do is understand that the MDC market has atypical needs and we do have solutions for this with LR4, 10X10-2km and 10X10-10km.
I suggest that we have a bifurcation in the market where we have what most people do and we have what the MDCs do. One piece of data on this is in Paul Kolesar's data where he shows over 5% of permanent links between patch panels in 2010 are over 119 meters long (see slide 8 of http://www.ieee802.org/3/100GNGOPTX/public/sept11/kolesar_02_0911_NG100GOPTX.pdf). Paul did not include these outliers in his analysis that shows only 1% of links are over 256 meters long (cell R31 of the Kolesar Kalculator). We need to understand what assumptions and limits we're putting on the data since this affects our conclusions.
I will present on this more in Newport Beach. I have gone to some large data centers, but not MDCs because they won't let people like me in! We need to make the distinction between the white space (the raised floor where equipment resides) and the building size of the data center. The white space is a significant subset of the building size. The white space can be separated by significant distances and this can lead to long permanent links between the white space.
I'm proposing that the nR4 solution would meet the needs of many links that are longer than SR4. From the Kolesar Kalculator, 19% of links are longer than 100 meters and less than 256 meters (cells R14-R31). 19% of links would need nR4 or LR4 if SR4 only supports 100 meters.
If SR4 matches the 150 meters of SR10 which will be very challenging with 25G lanes, then the nR4 market would shrink to about 8% of links (cells R19 to R31). Paul told me that the 2:1 mix is the most likely deployment scenario in the industry and corresponds to column R in the spreadsheet. Only 1% of links are longer than 256 meters according to Paul's work.
I hope that helps,
Scott
-----Original Message-----
From: Jeffery Maki [mailto:jmaki@xxxxxxxxxxx]
Sent: Friday, November 18, 2011 5:05 PM
To: Scott Kipp; STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx
Subject: RE: [802.3_100GNGOPTX] Emerging new reach space
Scott,
Was the choice to end your table at 400,000 sq. ft. arbitrary?
All,
I believe we need to know if the potential square footage may or may not grow larger over the coming years for what is known as a mega datacenter. How big is a mega datacenter to be? At some point, 100GBASE-LR4 will be the right choice just based on loss budget. We need to know the distribution of reaches to understand where to draw the line in selecting a break in the PMD definitions.
Jeff
-----Original Message-----
From: Scott Kipp [mailto:skipp@xxxxxxxxxxx]
Sent: Friday, November 18, 2011 1:38 PM
To: STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_100GNGOPTX] Emerging new reach space
Chris and all,
I have been wanting to discuss the reach objective for 100GBASE-nR4, so thanks for kicking off this discussion.
You referenced the 10X10 MSA white paper that calls out a maximum distance of <500 meters. You reference the authors of Vijay and Bikash, but I was the co-author that wrote this section of the paper and did the mathematical analysis which they agreed to. The actual distance of 414 meters is a simple calculation based on a 400,000 sq ft data center. Even if the data center is 550,000 sq ft, the link distance is less than 500 meters long. So I propose that 500 meters is long enough for the largest data centers that we should target.
The problem with a 500 meter distance is in the way that the IEEE defines the maximum link length. The IEEE defines the reach objective for SM fibers and gives an insertion loss based on the 2.0dB of connector and splice loss and the fiber attenuation loss. Specifically, 802.3ba states this below table 87.9:
The channel insertion loss is calculated using the maximum distance specified in Table 87–6 and cabled optical fiber
attenuation of 0.47 dB/km at 1264.5 nm plus an allocation for connection and splice loss given in 87.11.2.1.
For the 10km link of 100GBASE-LR4, the attenuation is 6.7dB = 10km * 0.47dB/km + 2.0dB of connector and splice loss.
If this project follows this example for a 500 meter nR4 link, then the insertion loss would only be 2.2dB = 0.5km * 0.47dB/km + 2.0dB for connector and splice loss. Many attendees know that this could limit the applicability of the nR4 link because it won't support structured cabling environments. With many MPO ribbon connectors in a link, it could be difficult to support a typical link in the structured cabling environments that will be required in large data centers.
To make nR4 a success, we need to take these structured cabling environments into account and increase the connector loss. I would like to hear from some cabling vendors and especially end users as to range of insertion losses that they have seen and what they expect to see if ribbon fibers are used instead of the usual duplex SM fibers.
Jonathan King did a great statistical analysis of 4 duplex MM connection loss in king_01_0508. We should do a similar analysis for 6 SM ribbon connectors to determine the loss of a long link.
If we determine the loss to be 3.5dB, then the insertion loss for the nR4 link could be 3.7dB = 0.5km * 0.47dB/km + 3.5dB for connector and splice loss. With these parallel solutions, this loss might even be larger since they don't have the WDM losses in the link.
That's my 42 cents,
Scott
-----Original Message-----
From: Chris Cole [mailto:chris.cole@xxxxxxxxxxx]
Sent: Friday, November 18, 2011 11:12 AM
To: STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_100GNGOPTX] Emerging new reach space
Jack,
Thank you for continuing to lead the discussion. I am hoping it encourages others to jump in with their perspectives, otherwise you will be stuck architecting the new standard by yourself with the rest of us sitting back and observing.
Your email is also a good prompt to start discussing the specific reach objective for 100GE-nR4. Since you mention 2000m reach multiple times in your email, can you give a single example of a 2000m Ethernet IDC link?
I am aware of many 150m to 600m links, with 800m mentioned as long term future proofing, so rounding up to 1000m is already conservative. I understand why several IDC operators have asked for 2km; it was the next closest existing standard reach above their 500m/600m need; see for example page 10 of Donn Lee's March 2007 presentation to the HSSG (http://www.ieee802.org/3/hssg/public/mar07/lee_01_0307.pdf). It is very clear what the need is, and why 2km is being brought up.
Another example of IDC needs is in a 10x10G MSA white paper (http://www.10x10msa.org/documents/10X10%20White%20Paper%20final.pdf), where Bikash Koley and Vijay Vusirikala of Google show that their largest data center requirements are met by a <500m reach interface.
In investigating the technology for 100GE-nR4, we may find as Pete Anslow has pointed out in NG 100G SG, that the incremental cost for going from 1000m to 2000m is negligible. We may then chose to increase the standardized reach. However to conclude today that this is in fact where the technology will end up is premature. We should state the reach objective to reflect the need, not our speculation about the capabilities of yet to be defined technology.
Thank you
Chris
-----Original Message-----
From: Jack Jewell [mailto:jack@xxxxxxxxxxxxxx]
Sent: Friday, November 18, 2011 9:38 AM
To: STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_100GNGOPTX] Emerging new reach space
Hello All,
Thanks for all the contributions to this discussion. Here's a synopsis and
my current take on where it's heading (all in the context of 150-2000m
links).
Starting Point: Need for significantly-lower cost/power links over
150-2000m reaches has been expressed for several years. Last week in
Atlanta, four technical presentations on the subject all dealt with
parallel SMF media. Straw polls of "like to hear more about ___" received
41, 48, 55, and 48 votes, the 41 for one additionally involving new fiber.
The poll "to encourage more on…duplex SMF PMDs" received 35 votes. Another
straw poll gave strong support for the most-aggressive low-cost target.
Impressions from discussion and Atlanta meeting: Systems users (especially
the largest ones) are strongly resistant to adopting parallel SMF. (not
addressing reasons for that position, just stating an observation.) LR4
platform can be extended over duplex SMF via WDM by at least one more
"factor-4" generation, and probably another (DWDM for latter); PAM and
line-rate increase may extend duplex-SMF's lifetime yet another
generation.
My Current Take: Given a 2-or-3-generation (factor-4; beyond 100GNGOPTX)
longevity of duplex SMF, I'm finding it harder to make a compelling case
for systems vendors to adopt parallel SMF for 100GNGOPTX. My current
expectation is that duplex SMF will be the interconnection medium. My
ongoing efforts will have more duplex-SMF content. I still think parallel
SMF should deliver lowest cost/power for 100GNGOPTX, and provide an
additional 1-2 generations of longevity; just don't see system vendors
ready to adopt it now.
BUT: What about the Starting Point (above), and the need for
significantly-lower cost/power?? If a compelling case is to be made for an
alternative to duplex SMF, it will require a very crisp and convincing
argument for significantly-lower cost/power than LR4 ("fair" comparison
such as mentioned earlier), or other duplex SMF approaches. Perhaps a
modified version of LR4 can be developed with lower-cost/power lasers that
doesn't reach 10km. If, for whatever reasons, systems vendors insist on
duplex SMF, but truly need significantly-lower cost/power, it may require
some compromise, e.g. "wavelength-shifted" SMF, or something else. Would
Si Photonics really satisfy the needs with no compromise? Without saying
they won't, it seems people aren't convinced, because we're having these
discussions.
Cheers, Jack
On 11/17/11 10:23 AM, "Arlon Martin" <amartin@xxxxxxxxxx> wrote:
Hello Jack,
To your first question, yes, we are very comfortable with LAN WDM
spacing. That never was a challenge for the technology. We have chosen to
perfect reflector gratings because of the combination of small size and
great performance. I am not sure exactly what you are asking in your
second question. There may be a slightly lower loss to AWGs than
reflector gratings. That difference has decreased as we have gained more
experience with gratings. For many applications like LR and mR, the much,
much smaller size (cost is related to size) of reflector gratings makes
them the best choice.
Thanks, Arlon
-----Original Message-----
From: Jack Jewell [mailto:jack@xxxxxxxxxxxxxx]
Sent: Thursday, November 17, 2011 6:42 AM
To: STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_100GNGOPTX] Emerging new reach space
Hi Arlon,
Thanks very much for this. You are right; I was referring to thin film
filters. My gut still tells me that greater tolerances should accompany
wider wavelength spacing. So I'm guessing that your manufacturing
tolerances are already "comfortable" at the LAN WDM spacing, and thus the
difference is negligible to you. Is that a fair statement? Same could be
true for thin film filters. At any rate, LAN WDM appears to have one
factor-4 generation advantage over CWDM in this discussion, and it's good
to hear of its cost effectiveness. Which brings up the next question. Your
data on slide 15 of Chris's presentation referenced in his message shows
lower insertion loss for your array waveguide (AWG) DWDM filter than for
the grating filters. Another factor-of-4 data throughput may be gained in
the future via DWDM.
Cheers, Jack
On 11/16/11 10:51 PM, "Arlon Martin" <amartin@xxxxxxxxxx> wrote:
Hello Jack,
As a maker of both LAN WDM and CWDM filters, I would like to comment on
the filter discussion. WDM filters can be thin film filters (to which you
may be referring) but more likely, they are PIC-based AWGs or PIC-based
reflector gratings. In our experience at Kotura with reflector gratings
made in silicon, both CWDM and LAN WDM filters work equally well and are
roughly the same size. It is practical to put 40 or more wavelengths on a
single chip. We have done so for other applications. There is plenty of
headroom for more channels when the need arises for 400 Gb/s or 1 Tbs.
There may be other reasons to select CWDM over LAN WDM, but, in our
experience, filters do not favor one choice over the other.
Arlon Martin, Kotura
-----Original Message-----
From: Jack Jewell [mailto:jack@xxxxxxxxxxxxxx]
Sent: Wednesday, November 16, 2011 9:09 PM
To: STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_100GNGOPTX] Emerging new reach space
Thanks Chris for your additions.
1. "CWDM leads to simpler optical filters versus "closer" WDM (LAN WDM)"
-
For a given throughput transmission and suppression of
adjacent-wavelength
signals (assuming use of same available optical filter materials), use of
a wider wavelength spacing can be accomplished with wider thickness
tolerance and usually with fewer layers. The wider thickness tolerance is
basic physics, with which I won't argue. In this context, I consider
"wider thickness tolerance" as "simpler."
2. "CWDM leads to lower cost versus "closer" WDM because cooling is
eliminated" - I stated no such thing, though it's a common perception.
Ali
Ghiasi suggested CWDM (implied by basing implementation on 40GBASE-LR4)
might be lower cost, without citing the cooling issue. Cost is a far more
complex issue than filter simplicity. You made excellent points regarding
costs in your presentation cited for point 1, and I cited LAN WDM
(100GBASE-LR4) advantages as "better-suited-for-integration, and
"clipping
off" the highest-temp performance requirement." We must recognize that at
1km vs 10km, chirp issues are considerably reduced.
3. "CWDM is lower power than "closer" WDM power" - I stated no such
thing,
though it's a common perception. I did say "More wavelengths per fiber
means more power per channel," which is an entirely different statement,
and it's darned hard to argue against the physics of it (assuming same
technological toolkit).
All I stated in the previous message are the advantages of CWDM (adopted
by 40GBASE-LR4) and LAN WDM (adopted by 100GBASE-LR4), without favoring
one over the other for 100GbE (remember we're talking ~1km, not 10km).
But
my forward-looking (crude) analysis of 400GbE and 1.6TbE clearly favors
LAN WDM over CWDM - e.g. "CWDM does not look attractive on duplex SMF
beyond 100GbE," whereas the wavelength range for 400GbE LAN 16WDM over
duplex SMF "is realistic." Quasi-technically speaking Chris, we're on the
same wavelength (pun obviously intended) :-)
Paul Kolesar stated the jist succinctly: "that parallel fiber
technologies
appear inevitable at some point in the evolution of single-mode
solutions.
So the question becomes a matter of when it is best to embrace them." [I
would replace "inevitable" with "desirable."] From a module standpoint,
it's easier, cheaper, lower-power to produce a x-parallel solution than a
x-WDM one (x is number of channels), and it's no surprise that last
week's
technical presentations (by 3 module vendors and 1 independent) had a
parallel-SMF commonality for 100GNGOPTX. There is a valid argument for
initial parallel SMF implementation, to be later supplanted by WDM,
particularly LAN WDM. With no fiber re-installations.
To very recent messages, we can choose which pain to feel first, parallel
fiber or PAM, but by 10TbE we're likely get both - in your face or
innuendo :-)
Jack
On 11/16/11 6:53 PM, "Chris Cole" <chris.cole@xxxxxxxxxxx> wrote:
Hello Jack,
You really are on a roll; lots of insightful perspectives.
Let me clarify a few of items so that they don't detract from your
broader ideas.
1. CWDM leads to simpler optical filters versus "closer" WDM (LAN WDM)
This claim may have had some validity in the past, however it has not
been the case for many years. This claim received a lot of attention in
802.3ba TF during the 100GE-LR4 grid debate. An example presentation is
http://www.ieee802.org/3/ba/public/mar08/cole_02_0308.pdf, where on
pages
13, 14, 15, and 16 multiple companies showed there is no practical
implementation difference between 20nm and 4.5nm spaced filters.
Further,
this has now been confirmed in practice with 4.5nm spaced LAN WDM
100GE-LR4 filters in TFF and Si technologies manufactured with no
significant cost difference versus 20nm spaced CWDM 40GE-LR4 filters.
If there is specific technical information to the contrary, it would be
helpful to see it as a presentation in NG 100G SG.
2. CWDM leads to lower cost versus "closer" WDM because cooling is
eliminated
This claim has some validity at lower rates like 1G or 2.5G, but is not
the case at 100G. This has been discussed at multiple 802.3 optical
track
meetings, including as recently as the last NG 100G SG meeting. We again
agreed that the cost of cooling is a fraction of a percent of the total
module cost. Even for a 40GE-LR4 module, the cost of cooling, if it had
to be added for some reason, would be insignificant. Page 4 of the above
cole_02_0308 presentation discusses why that is.
This claim to some extent defocuses from half a dozen other cost
contributors which are far more significant. Those should be at the top
of the list instead of cooling. Further, if cooling happens to enable a
technology which reduces by a lot a significant cost contributor, then
it
becomes a big plus instead of an insignificant minus.
If there is specific technical information to the contrary, a NG 100G SG
presentation would be a great way to introduce it.
3. CWDM is lower power than "closer" WDM power.
The real difference between CWDM and LAN DWDM is that un-cooled is lower
power. However how much lower strongly depends on the specific transmit
optics and operating conditions. In 100G module context it can be 10% to
30%. However, for some situations it could be a lot more savings, and
for
others even less. No general quantification of the total power savings
can be made; it has to be done on a case by case basis.
Chris
-----Original Message-----
From: Jack Jewell [mailto:jack@xxxxxxxxxxxxxx]
Sent: Wednesday, November 16, 2011 3:20 PM
To: STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_100GNGOPTX] Emerging new reach space
Great inputs! :-)
Yes, 40GBASE-LR4 is the first alternative to 100GBASE-LR4 that comes to
mind for duplex SMF. Which begs the question: why are they different?? I
can see advantages to either: (40G CWDM vs 100G closerWDM) - uncooled,
simple optical filters vs better-suited-for-integration, and "clipping"
off" the highest-temp performance requirement.
It's constructive to look forward, and try to avoid unpleasant surprises
of "future-proof" assumptions (think 802.3z and FDDI fiber - glad I
wasn't
there!). No one likes "forklift upgrades" except maybe forklift
operators,
who aren't well-represented here. Data centers are being built, so
here's
a chance to avoid short-sighted mistakes. How do we want 100GbE, 400GbE
and 1.6TbE to look (rough guesses at the next generations)? Here are 3
basic likely scenarios, assuming (hate to, but must) 25G electrical
interface and no electrical mux/demux. Considering duplex SMF,
4+4parallel
SMF, and 16+16parallel SMF:
Generation
100GbE duplex-SMF / 4WDM 4+4parallel / no WDM
16+16parallel / dark fibers
400GbE duplex-SMF / 16WDM 4+4parallel / 4WDM
16+16parallel / no WDM
1.6TbE duplex-SMF / 64WDM 4+4parallel / 16WDM
16+16parallel / 4WDM
The above is independent of distances in the 300+ meter range we're
considering. Yes, there are possibilities of PAM encoding and electrical
interface speed increases. Historically we've avoided the former, and
the
latter is expected to bring a factor of 2, at most, for these
generations.
Together, they might bring us forward 1 factor-of-4 generation further.
For 40GbE or 100GbE, 20nm-spaced CWDM is nice for 4WDM (4 wavelengths).
At
400GbE, 16WDM CWDM is a 1270-1590nm stretch, with 16 laser products
(ouch!). 20nm spacing is out of the question for 64WDM (1.6TbE). CWDM
does
not look attractive on duplex SMF beyond 100GbE.
OTOH, a 100GBASE-LR4 - based evolution on duplex SMF, with ~4.5nm
spacing,
is present at 100GbE. For 400GbE, it could include the same 4
wavelengths,
plus 4-below and 12-above - a 1277.5-1349.5nm wavelength span, which is
realistic. The number of "laser products" is fuzzy, as the same
epitaxial
structure and process (except grating spacing) may be used for maybe a
few, but nowhere near all, of the wavelengths. For 1.6TbE 64WDM, LR4's
4.5nm spacing implies a 288nm wavelength span and a plethora of "laser
products." Unattractive.
On a "4X / generational speed increase," 4+4parallel SMF gains one
generation over duplex SMF and 16+16parallel SMF gains 2 generations
over
duplex SMF. Other implementations, e.g. channel rate increase and/or
encoding, may provide another generation or two of "future
accommodation."
The larger the number of wavelengths that are multiplexed, the higher
the
loss budget that must be applied to the laser-to-detector (TPlaser to
TPdetector) link budget. More wavelengths per fiber means more power per
channel, i.e. more power/Gbps and larger faceplate area. While duplex
SMF
looks attractive to systems implementations, it entails significant(!!)
cost implications to laser/transceiver vendors, who may not be able to
bear "cost assumptions," and additional power requirements, which may
not
be tolerable for systems vendors.
I don't claim to "have the answer," rather attempt to frame the question
pointedly "How do we want to architect the next few generations of
Structured Data Center interconnects?" Insistence on duplex SMF works
for
this-and-maybe-next-generation, then may hit a wall. Installation of
parallel SMF provides a 1-or-2-generation-gap of "proofing," with higher
initial cost, but with lower power throughout, and pushing back the need
for those abominable "forklift upgrades."
Jack
On 11/16/11 1:00 PM, "Kolesar, Paul" <PKOLESAR@xxxxxxxxxxxxx> wrote:
Brad,
The fiber type mix in one of my contributions in September is all based
on cabling that is pre-terminated with MPO(MTP)array connectors.
Recall
that single-mode fiber represents about 10 to 15% of those channels.
Such cabling infrastructure provides the ability to support either
multiple 2-fiber or parallel applications by applying or removing
fan-outs from the ends of the cables at the patch panels. The fan-outs
transition the MPO terminated cables to collections of LC or SC
connectors. If fan-outs are not present, the cabling is ready to
support
parallel applications by using array equipment cords. As far as I am
aware this pre-terminated cabling approach is the primary way data
centers are built today, and has been in practice for many years. So
array terminations are commonly used on single-mode cabling
infrastructures. While that last statement is true, it could leave a
distorted impression if I also did not say that virtually the entire
existing infrastructure e!
mploys fan-outs today simply because parallel applications have not
been
deployed in significant numbers. But migration to parallel optic
interfaces is a matter of removing the existing fan-outs. This is what
I
tried to describe at the microphone during November's meeting.
Regards,
Paul
-----Original Message-----
From: Brad Booth [mailto:Brad_Booth@xxxxxxxx]
Sent: Wednesday, November 16, 2011 11:34 AM
To: STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_100GNGOPTX] Emerging new reach space
Anyone have any data on distribution of parallel vs duplex volume for
OM3/4 and OS1?
Is most SMF is duplex (or simplex) given the alignment requirements?
It would be nice to have a MMF version of 100G that doesn't require
parallel fibers, but we'd need to understand relative cost differences.
Thanks,
Brad
-----Original Message-----
From: Ali Ghiasi [aghiasi@xxxxxxxxxxxx<mailto:aghiasi@xxxxxxxxxxxx>]
Sent: Wednesday, November 16, 2011 11:04 AM Central Standard Time
To: STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_100GNGOPTX] Emerging new reach space
Jack
If there is another LR4 PMD out there the best starting point would be
40Gbase-LR4, look at its cost structure, and build a 40G/100G
compatible
PMD.
We also need to understand the cost difference between parallel MR4 vs
40Gbase-LR4 (CWDM). The 40Gbase-LR4 cost with time could be assumed
identical to the new 100G MR4 PMD. Having this baseline cost then we
can
compare its cost with 100GBase-LR4 and parallel MR4. The next step is
to
take
into account higher cable and connector cost associated with parallel
implementation then identify at what reach it gets to parity with 100G
(CWDM) or
100G (LAN-WDM).
In the mean time we need to get more direct feedback from end users if
the parallel SMF is even an acceptable solution for reaches of 500-1000
m.
Thanks,
Ali
On Nov 15, 2011, at 8:41 PM, Jack Jewell wrote:
Thanks for this input Chris.
I'm not "proposing" anything here, rather trying to frame the
challenge,
so that we become better aligned in how cost-aggressive we should be,
which guides the technical approach. As for names, "whatever works" :-)
It would be nice to have a (whatever)R4, be it nR4 or something else,
and
an english name to go with it. The Structured Data Center (SDC) links
you
describe in your Nov2011 presentation are what I am referencing, except
for the restriction to "duplex SMF." My input is based on use of any
interconnection medium that provides the overall lowest-cost,
lowest-power solution, including e.g. parallel SMF.
Cost comparisons are necessary, but I agree tend to be dicey. Present
10GbE costs are much better defined than projected 100GbE NextGen
costs,
but there's no getting around having to estimate NextGen costs, and
specifying the comparison. Before the straw poll, I got explicit
clarification that "LR4" did NOT include mux/demux IC's, and therefore
did not refer to what is built today. My assumption was a "fair" cost
comparison between LR4 and (let's call it)nR4 - at similar stage of
development and market maturity. A relevant stage is during delivery of
high volumes (prototype costs are of low relevance). This does NOT
imply
same volumes. It wouldn't be fair to project ER costs based on SR or
copper volumes. I'm guessing these assumptions are mainstream in this
group. That would make the 25% cost target very aggressive, and a 50%
cost target probably sufficient to justify an optimized solution. Power
requirements are a part of the total cost of ownership, and should be
consider!
ed, but perhaps weren't.
The kernel of this discussion is whether to pursue "optimized
solutions"
vs "restricted solutions." LR4 was specified through great scrutiny and
is expected to be a very successful solution for 10km reach over duplex
SMF. Interoperability with LR4 is obviously desirable, but would a
1km-spec'd-down version of LR4 provide sufficient cost/power savings
over
LR4 to justify a new PMD and product development? Is there another
duplex
SMF solution that would provide sufficient cost/power savings over LR4
to
justify a new PMD and product development? If so, why wouldn't it be
essentially a 1km-spec'd-down version of LR4? There is wide perception
that SDC's will require costs/powers much lower than are expected from
LR4, so much lower that it's solution is a major topic in HSSG. So far,
it looks to me like an optimized solution is probably warranted. But
I'm
not yet convinced of that, and don't see consensus on the issue in the
group, hence the discussion.
Cheers, Jack
From: Chris Cole
<chris.cole@xxxxxxxxxxx<mailto:chris.cole@xxxxxxxxxxx>>
Reply-To: Chris Cole
<chris.cole@xxxxxxxxxxx<mailto:chris.cole@xxxxxxxxxxx>>
Date: Tue, 15 Nov 2011 17:33:17 -0800
To:
<STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx<mailto:STDS-802-3-100GNGOPTX@L
I
S
T
SERV.IEEE.ORG>>
Subject: Re: [802.3_100GNGOPTX] Emerging new reach space
Hello Jack,
Nice historical perspective on the new reach space.
Do I interpret your email as proposing to call the new 150m to 1000m
standard 100GE-MR4? ☺
One of the problems in using today’s 100GE-LR4 cost as a comparison
metric for new optics is that there is at least an order of magnitude
variation in the perception of what that cost is. Given such a wide
disparity in perception, 25% can either be impressive or inadequate.
What I had proposed as reference baselines for making comparisons is
10GE-SR (VCSEL based TX), 10GE-LR (DFB laser based TX) and 10GE-ER (EML
based TX) bit/sec cost. This not only allows us to make objective
relative comparisons but also to decide if the technology is suitable
for
wide spread adoption by using rules of thumb like 10x the bandwidth
(i.e. 100G) at 4x the cost (i.e. 40% of 10GE-nR cost) at similar high
volumes.
Using these reference baselines, in order for the new reach space
optics
to be compelling, they must have a cost structure that is referenced to
a
fraction of 10GE-SR (VCSEL based) cost, NOT referenced to a fraction of
10GE-LR (DFB laser based) cost. Otherwise, the argument can be made
that
100GE-LR4 will get to a fraction of 10GE-LR cost, at similar volumes,
so
why propose something new.
Chris
From: Jack Jewell [mailto:jack@xxxxxxxxxxxxxx]
Sent: Tuesday, November 15, 2011 3:06 PM
To:
STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx<mailto:STDS-802-3-100GNGOPTX@LI
S
T
S
ERV.IEEE.ORG>
Subject: [802.3_100GNGOPTX] Emerging new reach space
Following last week's meetings, I think the following is relevant to
frame our discussions of satisfying data center needs for low-cost
low-power interconnections over reaches in the roughly 150-1000m range.
This is a "30,000ft view,"without getting overly specific.
Throughout GbE, 10GbE, 100GbE and into our discussions of 100GbE
NextGenOptics, there have been 3 distinct spaces, with solutions
optimized for each: Copper, MMF, and SMF. With increasing data rates,
both copper and MMF specs focused on maintaining minimal cost, and
their
reach lengths decreased. E.g. MMF reach was up to 550m in GbE, then
300m
in 10GbE (even shorter reach defined outside of IEEE), then 100-150m in
100GbE. MMF reach for 100GbE NextGenOptics will be even shorter unless
electronics like EQ or FEC are included. Concurrently, MMF solutions
have
become attractive over copper at shorter and shorter distances. Both
copper and MMF spaces have "literally" shrunk. In contrast, SMF
solutions
have maintained a 10km reach (not worrying about the initial 5km spec
in
GbE, or 40km solutions). To maintain the 10km reach, SMF solutions
evolved from FP lasers, to DFB lasers, to WDM with cooled DFB lasers.
The
10km solutions increasingly resemble longer-haul telecom solutions. T!
here is an increasing cost disparity between MMF and SMF solutions.
This
is an observation, not a questioning of the reasons behind these
trends.
The increasing cost disparity between MMF and SMF solutions is
accompanied by rapidly-growing data center needs for links longer than
MMF can accommodate, at costs less than 10km SMF can accommodate. This
has the appearance of the emergence of a new "reach space," which
warrants its own optimized solution. The emergence of the new reach
space
is the crux of this discussion.
Last week, a straw poll showed heavy support for "a PMD supporting a
500m
reach at 25% the cost of 100GBASE-LR4" (heavily favored over targets of
75% or 50% the cost of 100GBASE-LR4). By heavily favoring the most
aggressive low-cost target, this vote further supports the need for an
"optimized solution" for this reach space. By "optimized solution" I
mean
one which is free from constraints, e.g. interoperability with other
solutions. Though interoperability is desirable, an interoperable
solution is unlikely to achieve the cost target. In the 3 reach spaces
discussed so far, there is NO interoperability between copper/MMF,
MMF/SMF, or copper/SMF. Copper, MMF and SMF are optimized solutions. It
will likely take an optimized solution to satisfy this "mid-reach"
space
at the desired costs. To repeat: This has the appearance of the
emergence
of a new "reach space," which warrants its own optimized solution.
Since
the reach target lies between "short reach" and "long reach," "mid!
reach" is a reasonable term
Without discussing specific technical solutions, it is noteworthy that
all 4 technical presentations last week for this "mid-reach" space
involved parallel SMF, which would not interoperate with either
100GBASE-LR4, MMF, or copper. They would be optimized solutions, and
interest in their further work received the highest support in straw
polls. Given the high-density environment of datacenters, a solution
for
the mid-reach space would have most impact if its operating power was
sufficiently low to be implemented in a form factor compatible with MMF
and copper sockets.
Cheers, Jack
|