12:15 PM -- There's no doubt that Alcatel-Lucent (NYSE: ALU) is making significant headway in gaining traction for its 100Gbit/s transport equipment, having announced numerous deployments in recent months.
There's also no doubt that the vendor is an optical transport innovator and one of the leaders in the market. But when the company claims, as it did in its most recent 100Gbit/s customer win announcement in Austria, that it "led the way in the introduction of 100G, [and] remains the only vendor delivering this level of speed and capacity on existing networks," what does it mean exactly? (See H3G Deploys AlcaLu's 100G.)
Well, here's the answer, as supplied by Harald Himmer, head of the Alcatel-Lucent business in Austria, in an email to Light Reading.
Alcatel-Lucent's 100G solution provides the most efficient use of spectrum on existing networks, enabling optimum speed and capacity, as it eliminates the need for guard bands when deployed side by side with 10G and 40G wavelengths.
Alcatel-Lucent is the only vendor that has been commercially shipping a single carrier solution in volume with 45+ references and numerous publicly announced wins such as H3G, which are mostly network-wide upgrades to 100G, and not selective, highly-engineered routes as claimed in most competitors' announcements.
Alcatel-Lucent is a recognized leader in 100G research and 100G single-carrier coherent commercialization resulting in the (live, revenue-generating) deployment of 100G on any fiber, any route, any network.
All credit to Himmer for providing that detail, but is that the same as "the only vendor delivering this level of speed and capacity on existing networks"?
Any thoughts welcome on the message boards below.
In the meantime, here's a rundown of AlcaLu's most recent 100Gbit/s news:
Cisco has alway been pretty good at announcing major wins. I will believe it when I see a series of wins at more than one national scale carrier. globally.
not disputing Cisco has a play. And I might be behind the curve. But I would be very surprised if they had that large a share in long haul transport considering I don't see the big announcments for network wins like we see everyone else do. I see wins announced by lots of companies, but news for Cisco major transport wins is a little thin. Again, maybe we will see this over the coming quarters.
Sorry I cannot show you the reports on market share on long distance DWDM transport for obvious reasons but you would be surprised to see that cisco is very well placed and growing while others have serious problems. If you can get the various reports check and you will understand what I am saying.
In term of 100 G you will see this year the growth in cisco share. May be in one year time you will agree with me.
Yeah, I sure wish I had known Chris was unhappy there. I would have hired him immediately! lol, he is one of the few in this field I would even give a big signing bonus too like they used to do in the old days during the dwdm boom of the late 90's!
Sure makes me curious about what they are doing at Acacia! I know he probably had his pick of opportunities.
Regards my point. Yes, there are several systems vendors. But if you look at it, when the market got dis-aggregated and deconstructed in the early 2000's, and everyone started bying the same merchant modules, transponders, transcievers, pluggables, etc. etc. it became like a micro soft / intel pc situation. in pc's yes there are also several providers. but the distinctions between them are almost zero. The only way you can distinguish is with software and services riding on top. a pc is a pc is a pc is apc.
With optical systems OEM's there has been increasing whispering about this problem in strategy meetings over the last few years. How do we differntiate our platforms?
Pric?: no: Huawei wins.
Until someone comes along and dethrones Huawei with a lower cost of labor (Bangladesh could be the next great mecca for low cost labor folks! All you need to do is make yoru factor on a barge that floats for when the floods come...hahaha)
Performance? oops, we all use the same modules. Only differentiation is when we do our redesign cycles so we intercept components suppliers at diffferent calendar windows.
software and NMS designs: ok, maybe. But is this enough?? and isn't this exactly what makes pc's different? the software and services that run on them are all different.
And didnt' the pc industry really fall into itself with now only really a few players left and a couple of national brands. (japan will always make their own)
So the model drives huge revneues to a very few in that model. Intel and Microsoft and a thin slice of surving integrators (now you are in integrator, not a system designer...after all, you are buying the same stuff) makes a living but not great returns. Except for Apple, how happy are you with stock market returns for pc companies over the last few years. They are all struggling trying to find ways to grow and differentiate.
Services? Worked for IBM and that is the conventional wisdom. But is that enough to save ALU, NSN, ERIC, etc. etc. etc. who are all being slowly ground down by the chinese invasion of high tech. If all buy the same modules, it is hard to differentiate. Huawei loves this. As does ZTE. This is not a judgement. I think both are good companies, both have made a lot of good jobs in china, and both have taken advantage of market conditions to leverage their strong points and minimze their weaknesses. I think this is wise and good manament. But it has played out just the opposite for the big players. Both at the systems level and at the modules level, margins are thin and it is getting tougher to survive. Take a look at the stocks in our sector. especially if you take a longer look at track the trends for say the last 30 years.
When you all buy the same transponders, same roadms, same transcivers, same amplifiers, the sytems begin to look more and more alike. Yes, there are some architecture things, but those are pretty limited. They all buy from the same back plane vendors, from the same power supply makers, same chassis makers, etc. So when optical modules are forced into this model you indeed get less innovation.
As you mentioned. the module guys have margins that are too thin to do much real reserach and development. And if the systems guys are not doing it, then who is?
You are misinterpreting the infonetics information.
Cisco did have a good upwared moving quarter based on new blades and modules. The link you pointed me to states very clearly this is Cisco growth over previous quarter. Now, one of us would have to publish the details of that infonetics report to have real info, and I am not going to do it. BUT, in the cover page from the link in no place does it talk about Cisco market share. Nor does it state what kind of modules caused this growth. Cisco has a big installed base in metro networks as for a long time they ran as one of the top 4 suppliers of metro DWDM Cisco also has a lot of installed Cerent boxes that later had a metro dwdm shelf added to them. And the new higher density optical modules on the routers. And they are starting to see some of the benefits of having their own high end transponder team in the CoreOptics acquistion.
But long haul transport? Cisco abandoned that long ago. Yes, they got their own transponder team by buying CoreOptics. But that lets them build nice line cards with differentiators based on the MLSE that Core designed. NOTE this MSLE engine could also be easliy modified to work on distortion issues for television signals. Cisco is still a player in TV delivery via the old Scientific acquisition. So they got their own internal transponder design team that will give them better line cards. But that is not a long haul transport high density end to end DWDM system.
I think you missed my distinction. I was talking about long haul transport. not metro, access, older sonet aggregation boxes, nor was I talking about upgrades on native optical modiules for router line cards. Those are the areas were CISCO has grown.
And again, there is no mention of market share in the infonetics cover. Cisco's revenue growth does not necessarily mean share growth. I think this is a blip based on the release of new line cards and upgrade cycles. And this is one of the first reports that captures the addition of Core Optics to Cisco's optical revenue.
It would be interesting to look inside the report, and also get a Del'oro report and an OVUm... then put it all into a spreadsheet.
I think the picture would be interesting and back up my points.
Tell me when was the last time Cisco designed or deployed a Cisco Long Haul EFDA or RAMAN system?
There are many other public informations that show the impressive growth of cisco in the transport system.
The real issue in 100 Gbit/s upgrading of existing line is as you know the presence of dispersion compensation which increase non linear effects and limit the reach. The winner will be the supplier who will show the better tollerance to dispersion compensation. No customer will accept to go in the field to refurbish the amplifier site.
'If so, innovation slows, we end up with a microsoft model were the only differentiator is price. Not performance. And if so, then folks like Finisar, Opnext, Oclaro, et al win. But less innovation. Only packaging of modules.'
I do not understand your comparison. Unlike the microsoft model, there are multiple companies generating designs. There is a lot of room for these companies to differentiate based upon performance. However, there is the elephant in the room, that module companies typically have lower margins and lower ASPs such that they have less wherewithal to innovate. Is this the point you were making? The flip side of that argument is that the module companies have more physical layer expertise in details like packaging and very high speed serial traces. There are some very good heads at the various systems companies mentioned in the thread, but the divestiture of components has left the module companies with a Marxist cliche 'Quantity has a quality all its own'.
Also, ALU and crew are going to miss Chris Doerr quite a bit in my opinion.
How much market share do you think they have in DWDM transport long haul????
I know Cisco keeps trying. But sorry, no cigar here folks.
The carriers have been burned too many times by Cisco "being in optical"... "phasing away from optical"... buying platforms and hyping them (shall I make a list??), then canceling them and dropping the platforms and teams.
Carries might put native transponder interfaces on Cisco routers that can slot into someone elses transport systems.... but Cisco being a real carrier LONG HAUL DWDM transport player?
and yes, NSN has a nice solution. For sure. Different than others as I said. Each of the big players rolled their own line cards for first releases of 100G and each did it differently.
There are often many factors going into a purchase decision for a carrier for optical transport. Span type? fiber type? installed base of EDFA's? is there RAMAN in the span? how many repeaters are needed? how far apart are the huts? And INCUMBENCY. Sure easier to add a line card to an existing system than install an entirely new system. And the big one: saves a BUNDLE on training of the technical staff inside a carriers if the solution is a new line card for an existing platform on which the technical staff has already been trained. This by the way is an often overlooked cost that many systems OEMs never think about to their detriment. I have won many deals over the years by understanding this and hence, when I have wanted to introduce new technology, I have found ways to mitigate this cost (and complexity and time) and make it easier to adopt.
Sailboat I agree on your comment on Ciena but NSN and Cisco have single wavelength operation just as Alcatel but obviously they have something more sophisticated because their reach is since the beginning 2500 km.
Indeed. But to be clear, if you are considering merits of technology, this is not an "apples to apples" comparison.
the Nortel (Ciena) solution uses 2 lambdas closely packed together. Each lambda carrying 50G. And each of those lambda's use modulation and detection techniques that all the actual electronics on each end of the optics to run slower.
ALU does it on one lambda without guard bands. Easier to get reach with 2 lambda's.
I think both are good solutions. Which is better? both have their places in the market.
I think that is a case by case decision the operator has to make based on the fiber route, what time of fiber, distance between huts, NMS platforms, etc.
And the second question of is this going to be an entirely new platform deployment, or is it an upgrade to the installed base. Incumbency means something in telco.
What will be interesting to watch is does the industry really coalesce around standard transponders and give up doing their own line card design? If so, innovation slows, we end up with a microsoft model were the only differentiator is price. Not performance. And if so, then folks like Finisar, Opnext, Oclaro, et al win. But less innovation. Only packaging of modules.
Or, will each systems OEM continue to pursue their own path seeking performance advantages and driving innovation?
It will be interesting to watch this unfold as 2nd and 3rd generation platforms get introduced.
For now, I see the big players all rolling their own line cards for 100G and the middle tier players buying transponders.
Each of the big guys all have a somewhat unique approach at this point. Despite what the transponder vendors would wish.
The blogs and comments are the opinions only of the writers and do not reflect the views of Light Reading. They are no substitute for your own research and should not be relied upon for trading or any other purpose.
More EuroBlog
SDN Added to Interop Test 11:00 AM OpenFlow and Path Computation Element (PCE) will be part of EANTC's annual multivendor Carrier Ethernet/MPLS interoperability test
To save this item to your list of favorite Light Reading content so you can find it later in your Profile page, click the "Save It" button next to the item.
