x
DWDM

Ceyba Goes Ultra-Long

The ultra-long-haul optical transport market is heating up. While the boom may still be at least a year out, well funded startups are gearing up to introduce the next generation of products.

Last week, Ottawa-based Innovance Networks announced it had raised $55 million (see Innovance Scores $55M). This week, crosstown rival Ceyba Inc. introduced its first product, the C420 Dynamic Optical Networking System, an ultra-long-haul dense wavelength-division multiplexing (DWDM) platform (see Ceyba Targets Core).

Like other ultra-long-haul transport products, the C420 will be able to transmit at distances up to 4,000 kilometers without the need for optical regeneration, which greatly reduces the need for OEO switching in the core. The company claims its platform can save carriers as much as 70 percent in capital costs.

But Ceyba says there are a few differentiators that set it apart from the rest of the pack. For one, Ceyba claims that it will be the first company to deliver an optical transport product that mixes 10-Gbit/s and 40-Gbit/s channels in the same transmission band, allowing carriers more flexibility in how they deploy the product. The 10-Gbit/s interfaces will be available on the product when it ships in June, while 40-Gbit/s interfaces will be available later.

“We are the only ones who will have a product ready for 40-Gbit/s from the start,” says Benoit Fleury, vice president of product marketing at Ceyba. “Corvis realized that 10-Gbit/s was a good idea, but I haven’t heard of any plans to go to 40-Gbit/s. We’ve future-proofed our platform.” Ceyba’s second big differentiator is its monitoring and diagnostic capabilities. Fleury says the product’s diagnostic tool kit is able to monitor bit error rate, forward error correction, and optical time domain reflection. It also includes an optical spectrum analyzer, along with other important measurement tools, all in one testing device. Fleury contends that ultra-long-haul transport technology hasn’t yet taken off in a big way partly because currently available systems lack these attributes.

These differentiators are all very well, but the market reality is that Ceyba is entering a crowded vendor pool in the middle of an economic recession. Not only is this market flooded with competition from incumbent players like Nortel Networks Corp. (NYSE/Toronto: NT) and Ciena Corp. (Nasdaq: CIEN), which sell traditional long-haul transport gear; but Ceyba also faces competition from ultra-long-haul players like Corvis Corp. (Nasdaq: CORV) and other well funded startups, including Innovance and PhotonEx Corp.

To make matters worse, the market for long haul, as well as ultra long haul, seems dead in the water. The industry is coming off a two- to three-year period of massive network buildouts that left many carriers with a capacity glut. As a result, carriers have drastically reduced their spending on long-haul gear. For example, both Qwest Communications International Inc. (NYSE: Q) and Sprint Corp. (NYSE: FON) have practically frozen all spending on their long-haul networks (see Qwest Slowdown Spooks Investors). Williams Communications Group (NYSE: WCG), a carrier that spent millions on long-haul gear during the boom -- and is now contemplating bankruptcy -- is expected to cut spending by at least another 25 percent (see Williams Ponders Bankruptcy).

All players in this market, including the big ones, Nortel and Ciena, have felt the pain (see Ciena: Outlook Dim). Corvis, the only company that has actually begun shipping ultra-long-haul gear, reported revenues of only $15.2 million last quarter (see Corvis: How Low Can It Go?). And revenues are expected to be flat to down for at least the next two quarters.

Fleury suggests that Corvis’s concept hasn't failed: It was merely too far ahead of the market.

“Their timing was off,” he says. “They came out with a product as the market was collapsing, but we have a fairly large window to deploy our product.”

It could be argued that Ceyba and other startups -- like Innovance, which now has raised over $130 million, and Photonex, which has over $178 million -- will actually benefit from the market slowdown.

“If you have cash to last you for the next couple of years, the downturn actually buys you some more time to get customer traction,” says Rick Schafer an analyst with CIBC World Markets.

Ceyba has raised over $93 million in funding and expects to have its product generally available for customer shipments in June of this year. But company executives say they don’t really expect to ship products in significant volume until at least the end of 2002 or the beginning of 2003.

According to Fleury, most carriers are only using about 70 percent of their lit fiber capacity. But traffic is still growing at about 100 percent per year, which means that in another year, these carriers will need to turn up capacity on more of their fibers.

“On average, you have to figure they will be exhausting lit fiber capacity in the next 12 months,” he says. “This gives us a chance to get our feet wet and start small -- then we can flesh out customer deployments.”

— Marguerite Reardon, Senior Editor, Light Reading
http://www.lightreading.com
Page 1 / 5   >   >>
opticalwatcher 12/4/2012 | 10:51:13 PM
re: Ceyba Goes Ultra-Long There's a statement in this article indicating that traffic growth is occuring at 100% per year. Does anyone know or have comments on how growth is measured and whether this growth rate is realistic and accurate?
thx
PhotonGolf 12/4/2012 | 10:51:13 PM
re: Ceyba Goes Ultra-Long
"According to Fleury, most carriers are only using about 70 percent of their lit fiber capacity. But traffic is still growing at about 100 percent per year, which means that in another year, these carriers will need to turn up capacity on more of their fibers."
____________________

Hmmm. So when they need to light more fiber, won't they just buy more linecards from their existing vendors? I guess when they fill to capacity, they'll need new systems.

Anyone have a number on the average percentage utilization of the existing long haul platforms? Could indicate just when new systems like Ceyba could see market opportunity.

skeptic 12/4/2012 | 10:51:12 PM
re: Ceyba Goes Ultra-Long There's a statement in this article indicating that traffic growth is occuring at 100% per year. Does anyone know or have comments on how growth is measured and whether this growth rate is realistic and accurate?
----------------

Some traffic is still growing at extremely
agressive rates, but even if the rates are 100%
per year, its going to take a long time to
make a dent in the existing long-haul capacity.
There are lots of fibers and lots of wavelengths
that could go on each fiber. Its just hard to
imagine anyone even thinking about long-haul
buildout again for a long while.



mrcasual 12/4/2012 | 10:51:11 PM
re: Ceyba Goes Ultra-Long You install a span of 100 fiber cable equipped with 32 lambda, 100 GHz
transmission gears and the appropriate amplifiers. You lit 10 lambdas
with 2.5 Gbps SONET. What percentage of your route is utilized?


Not very much. Do I get a passing grade?

The issue with increasing traffic wrt to long haul buildouts isn't as much the fiber in the ground as it is the terminating equipment.

Going forward I think gear that connects to the long haul part of the network will continue to sell as traffic grows. How much of that is new boxes, or upgrades to existing boxes is the really tough question for a lot of people right now.

If Ceyba (or others) are required to connect that new traffic then they have a chance. If the incumbants can address the expansion with relatively painless (from both and opex and capex POV) upgrades then they are doomed.

Note that it's not only a "transmission" problem. New routers, ATM switches (blasphemy!), etc. will all be required as traffic grows so there is hope assuming traffic continues to grow at a reasonably high rate, which all studies seem to conclude.
Milano 12/4/2012 | 10:51:11 PM
re: Ceyba Goes Ultra-Long Utilization rate is meaningless and indeed mislead quite a few people last year. Try to answer the following question:

You install a span of 100 fiber cable equipped with 32 lambda, 100 GHz transmission gears and the appropriate amplifiers. You lit 10 lambdas with 2.5 Gbps SONET. What percentage of your route is utilized?

I bet you will have as many answers as there are ID's on LR...

M.
melao 12/4/2012 | 10:51:10 PM
re: Ceyba Goes Ultra-Long "The issue with increasing traffic wrt to long haul buildouts isn't as much the fiber in the ground as it is the terminating equipment.

Going forward I think gear that connects to the long haul part of the network will continue to sell as traffic grows. How much of that is new boxes, or upgrades to existing boxes is the really tough question for a lot of people right now.

If Ceyba (or others) are required to connect that new traffic then they have a chance. If the incumbants can address the expansion with relatively painless (from both and opex and capex POV) upgrades then they are doomed.

Note that it's not only a "transmission" problem. New routers, ATM switches (blasphemy!), etc. will all be required as traffic grows so there is hope assuming traffic continues to grow at a reasonably high rate, which all studies seem to conclude."

That-¦s the whole point. We have several DWDM long-haul systems with 32 or more wavelengths that has only 4 to 10 wavelengths occupied. Even if the traffic grows A LOT it-¦s just a matter of adding line cards.

I find very weird to bet on long-haul for a while.
And ultra-long-haul. Geez, how many carriers need ultra-long-haul nowadays ?

My guess is that in the beginning of 2004 we-¦ll see some substantial growth on long-haul... Just a guess.
opto 12/4/2012 | 10:51:09 PM
re: Ceyba Goes Ultra-Long Various sources are estimating traffic growth at 80% annual to 120%. Definitions of "traffic" are non-trivial, and no one agrees what constitutes the "proper" measure of network traffic, so even these numbers are really only rough numbers. The overall growth looks to be continuing at about the same pace over the last 5 years, minus the blip in 2K over dot.com mania.

I have some circumstantial evidence that typical fills of lambdas are on the order of 10% in LH DWDM. My numbers are from only one (big) carrier. Also, this would explain Huber saying Corvis does not need lamda conversion. With very low fill rates, it is simply not needed. He did not say it would "never" be needed, did he???

The issue of whether other dwdm systems vendors can make it is somewhat less clear, even though fill rates are low. I have heard of several vendors recently installed gear being replaced by another vendor's gear because the operations expenses associated with adding lambdas turned out to be too high. So that gear never got more than the initial lambdas lit and it was trashed (read second-hand market).

The fundamental issue facing carriers today is this: Their constituency (bosses) has changed their objective. In the 90's, it was "grow at all costs". Now, it is "find a way to be profitable". Wall Street has changed it's expectatations. Carriers are working to adapt. They must find lower cost ways to operate their networks. So now, Opex, or Total Cost of Ownership, is a big driver. It never was in the 90's. Thus, if a new vendor comes to market with a system that costs less to install, provision, maintain, monitor, troubleshoot, and expand, well then, they just might have a chance. Actually, they might have a really good chance. But they have to hope that the incumbents have designed themselves into corners, and that they cannot improve features and performance to the point that the new vendors do not have an Opex advantage.

So, no gross generalizations work here. Can Ceyba, et al, survive? They have a chance, if the gear proves in.
secretIdentity 12/4/2012 | 10:51:05 PM
re: Ceyba Goes Ultra-Long You're probably right with respect to expanding the number of wavelengths on a existing fiber - I think that might be tough to do with equipment from multiple vendors, and still yield a cost savings.

But what about expansion to light up the next fiber? I think using different vendors on different fibers in the same bundle would be OK.

And if a new vendor can save the carrier capex and opex with new equipment, how do you justify using the same old (more expensive) equipment?

I don't think Ceyba is providing the same old line cards at 70% off. They've probably got something significantly different to offer. If they don't, and the incumbent equipment vendors can just chop 70% off their prices with cost-reduction exercises, then why haven't they done this already?

However, I think the 100%/yr. growth rate sounds a little dated. If todays (or yesterdays) equipmemt was so expensive, how could anyone afford to overbuild by 100's of %? But I could be wrong on this one, too.
citpo 12/4/2012 | 10:51:03 PM
re: Ceyba Goes Ultra-Long Innovance, Ceyba, Covis
Any comment about their strength and weakness?
Fhunton 12/4/2012 | 10:51:02 PM
re: Ceyba Goes Ultra-Long Just reading some of the comments here, and it is exactly what i feel is wrong in the industry. There is lots of un-utilised lambdas. I work for a major telco vendor in Network Integration for DWDM.

We use our DWDM product in all the main operators networks, this machine obviously has massive capacity, carries over 80 lambdas (base model) etc etc.....the carriers are probably only using on average 8-10 lambdas!!!. The biggest one i've seen is 18 lambdas, and these are major carriers, and our company is launching it's next gen long haul ultra capacity product, when it's base product hasn't been fully utilised yet, I know and realise there are other factors by introducing new products etc.....but what i'm basically saying is just reiterating what has been said already in the this article, there are alot of unused lambdas that are going to take along time to fill......how much capacity do we need, i think it's alot less than is perceived, and is being touted around......utlilisation of existing network infrastructure is the next direction for telecoms, i personally think......
Page 1 / 5   >   >>
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE