Briefing.com Photonics Tutorial, Part Deux
tekboy/Ares@forpartone,see25229above.com
Photonic Revolution II: DWDM
[BRIEFING.COM - Gregory A. Jones] In the first part of our series on optical networking, we discussed the sector in the broadest of terms -- comparing it to the transportation network. In this second part, we begin to look at specific fiber optic sub-sectors. As we travel through the fiber network, we'll start once again at the core of the network, where transoceanic and transcontinental cables have one mission -- transmit huge amounts of voice and data over long distances. And the place to start is with the technology that truly unleashed the power of fiber: DWDM.
The Technology
DWDM is dense wave division multiplexing. Prior to DWDM, the capacity of an individual strand of fiber was the capacity of the one channel on that strand. That bandwidth is already considerable and growing larger, with current OC-192 fiber delivering 10 gbps and next generation OC-768 fiber delivering 40 gbps (the OC numbers ramp in tandem with transmission capacity; ie a quadrupling of the OC # from 192 to 768 is synonymous with a quadrupling of capacity).
But DWDM greatly increased fiber's capacity by allowing for multiple channels on one strand of fiber. These individual channels represent different wavelengths, each wavelength having the same potential bandwidth. So an OC-192 fiber strand with 10 gpbs capacity can be increased to 400 gpbs capacity with a 40-channel DWDM system.
It is this technology that drives the scalability of fiber networks discussed in Robert V. Green's brief yesterday. It is extremely expensive to bury a conduit containing fiber strands, but once it is buried, it is relatively easy to "blow" new strands through the conduit and easier still to add channels to each strand by employing DWDM technology. For the core of the telecommunications network, also known as long-haul transmission, DWDM is an obvious choice to meet soaring bandwidth demand.
The Players
With exploding data traffic driving bandwidth demand, it is hardly surprising that sales of DWDM systems are also booming. Pioneer Consulting projects that the market for DWDM systems will soar to $17.9 bln in 2004 from $5.6 bln this year. The vast majority of these sales will go to the long haul market, which is dominated by a few major players. The remaining sales will go to the nascent metro DWDM market, which will see larger growth rates but much smaller numbers. Today we will profile only the players in the long haul market; we will return in a future brief to the metro DWDM players.
Nortel (NT) owns the largest share of the long haul DWDM market. Pioneer estimates its share at 33%, while a recent Dell'Oro study placed its share at a more dominant 45.5%. Nortel's high-end DWDM system delivers up to 320 gbps with 32 channel DWDM on OC-192 fiber.
Lucent (LU) is acknowledged as the number two player in the industry with 27% market share, according to Pioneer. Its WaveStar OLS 400G optical networking system delivers up to 400 gbps capacity with 40 OC-192 channels.
CIENA (CIEN) has about 15% market share. Its MultiWave CoreStream is capable of 480 gpbs on 48 OC-192 channels.
Alcatel (ALA) is right alongside CIEN at 14% share. Its OPTINEXTM 1640 WM delivers 40 gbps on 40 OC-192 channels.
As you venture below these top four, the market shares fall off considerably. Some of the remaining players are Cisco (CSCO) via its purchase of Pirelli's DWDM division, NEC, and Ericcson (ERICY), all in the 3-5% market share area.
Caveats
Several points are important to note when you consider both the market shares and DWDM capabilities of the companies mentioned above. On the market share front, it is important not to look at the DWDM business of NT, LU, and ALA and conclude that these are explosive companies whose stocks are bound to match the rise in the DWDM market. For all three of these companies, DWDM systems comprise a very small chunk of sales. Even optical systems in their entirety are a relatively small portion of sales. We will discuss these matters more as this series progresses.
Only CIEN is close to a pure play. It is a pure play on optical networking, but not on DWDM specifically. It has added to its product line recently, and its future depends importantly on the success of its optical switch and long haul transmission products, which -- sorry, once again -- we will write about in a later brief.
As for the technologies, you must keep in mind that it is difficult to stay on top of which company has the superior DWDM product. Not only does the technology change very fast, but companies have a bad habit of preannouncing products that don't yet exist, as the promise of a great future product can make it more difficult for a competitor to sell its less attractive, but real, current products.
Lucent is already talking about a system that will produce 800 gbps bandwidth early in 2001, and its Bell Labs arm has announced a successful test of a 3.28 terabit (tbps) system. NEC has also said that it will have a 3.2 tbps system available in 2001 with an 80 channel OC-768 product.
While we will almost certainly see these bandwidth thresholds met and surpassed, preannouncements make it difficult to know who is actually selling what and when. Don't assume that multi-terabit announcements mean that any of these companies has jumped to a huge technological advantage -- go take a look at what they are actually selling in the here and now.
The Complete Package
What you will find is that current products are quite comparable, and that competition takes place primarily on price and the ability to offer a complete solution. That latter factor is what is driving the acquisition frenzy in the optical space. No one currently offers a complete optical solution, because that complete solution doesn't yet exist. But everyone is out to acquire its way to that ultimate solution.
Lucent bought Chromatis. Cisco bought Cerent, Monterey, Qeyton, and Pirelli's DWDM division. Nortel bought Xros, Qtera, and CoreTek. Ciena bought Lightera. And now our favorite optical site -- Light Reading -- is reporting that Sycamore (SCMR) in on the prowl for a metro optical firm (names mentioned are all private -- Appian, Alidian, Astral Point, Mayan, and Sirocco).
As we negotiate this rapidly shifting sector, we will try to determine who is best positioned to benefit from the shift to the all-optical network. Winning technologies and broad product offerings will be key. Today we got a quick picture of the long-haul DWDM players, a key component of that total optical package. We will continue to move down through the network as the series progresses, discussing other technologies key to the core of the network and moving into the metro area and finally the edge. Stay tuned...
Parting Shots
* Photonic Clarification: More than one astute Briefing.com reader noted after our first piece that the key to fiber's ascendancy is not so much that photons travel faster than electrons -- they only travel a little bit faster. The key is that photons have no mass and thus much higher possible frequencies which allow photons to carry more information.
* Supercomm2000: Expect a slew of product announcements and perhaps even a few more acquisitions as the year's largest telecom equipment conference gets underway in Atlanta from June 4-8. As we noted above, read the product announcements skeptically as most of them will contain more future promise than present greatness. |
