Tunable-laser market down but not out
There's no denying that the tunable-laser market is in the midst of consolidation. ADC Telecommunications (Minneapolis) closed its tunable-laser facility in Sweden and later announced that it was looking to sell its entire optical components division. Nortel Networks (Ottawa, Ontario) is also reportedly looking for a buyer for its optical components business. And New Focus (San Jose, CA) announced in May the selling of its network tunable-laser technology to Intel (Santa Clara, CA). With deployments pushed back to the second half of 2003 and even into 2004, vendors are reassessing their business strategies.
For those who decide to weather the storm, the near-term outlook is bleak but not hopeless. Reported design wins and Telcordia qualification announcements indicate that the market is moving forward, however slowly. Says one industry insider, "Deployment may have been delayed, but the market certainly isn't dead."
The narrowly tunable-laser market, while by no means robust, is at least showing signs of life, asserts Tom Hausken, director of optical communication components at market researcher Strategies Unlimited (Mountain View, CA). "Something is happening in the tunable-laser market, and it's happening in this narrowly tunable range," he says, "which is either anti-climatic or it's big news, depending on your view."
Narrowly tunable lasers are "not much of a modification over today's fixed wavelength technology, because today's lasers have temperature control," explains Hausken. "What you do is add a feature that allows you to change that temperature, and now you have a tunable laser over a small range—which is good enough for some systems people right now."
Furukawa Electric Co. (Tokyo) recently unveiled a 20-mW tunable distributed-feedback (DFB) laser for long-haul and metro DWDM applications, while Alcatel (Paris) and Fujitsu Network Communications (Richardson, TX) have incorporated narrowly tunable devices into their own DWDM transmission systems. Alcatel uses a two-channel tunable laser in its 1696 MetroSpan Release 1.0 DWDM platform, and Fujitsu has incorporated a four-channel thermally tuned DFB in its Flashwave systems.
Narrowly tunable lasers are low-cost devices, mainly used now as replacements for fixed wavelength lasers. It is their widely tunable counterparts that have the potential to be revolutionary.
There are as many different tunable technologies as there are companies pursuing them, and by all accounts, it's still too early to predict a clear winner. The devices still suffer from reliability concerns, and cost remains an issue.
Of course, tunable-laser vendors are quick to point out that systems vendors and carriers still have tunability on their roadmaps. So the tricky question is not so much "will they" but "when." There's no ready answer, says Saeid Aramdeh, vice president of product management at iolon (San Jose).
"The top 10 carriers in North America account for about 70-80% of long-haul spending," he reports. "Out of that list, about four are already under Chapter 11. And the market in Europe is not much better, with Flag and Carrier 1 both closing their doors, followed by KNPQwest." This global reduction in capital expenditures has all but ceased new builds, which in turn has seriously impacted the long-haul market.
That said, some industry insiders believe the tunable-laser market would have stalled anyway, regardless of market conditions. "To tell you the truth, even if things were going well, I don't think these companies would be doing that much better," contends Hausken. "The industry kind of got ahead of itself. Companies have all these new-fangled ideas, and then finally it comes down to putting these things into systems and making those systems work, and it turns out that companies don't really want to go that fast—even in good times. My feeling is that the tunable-laser market was a bit overblown even before the downturn, but of course, the downturn makes it worse," he adds.
The more critical question therefore seems to be not "when will tunable lasers be ready" but "when will the application space be ready for tunable lasers?" The devices have long been championed as the enabler of dynamic provisioning, but this application has taken a back seat to the more immediate need for inventory management.
Today, narrowly tunable lasers are deployed primarily for inventory management, which requires a low-cost laser solution. They have also been used for one-time provisioning, in which the laser is set to one channel for the duration of its life, but these applications do not take advantage of the full benefit of tunability, notes iolon's Aramideh. "Carriers are not asking for tunability just for sparing," he says. "They are asking for the full value of tunability as it comes to bear on provisioning and bandwidth management."
"Dynamic provisioning is there, waiting," agrees Hausken. "It's just going to take some time. Nobody is buying systems. And nobody knows which architecture to use. But it's there."
The general consensus seems to be that deployment of widely tunable lasers will begin the second half of 2003 or early 2004. That does not mean 2002 will end up a total wash, however. On the contrary, it may prove to be a critical year for design wins. Agility Communications (Santa Barbara, CA) has reportedly nabbed two major tier one design wins and, at press time, was in the running for two more. Iolon claims to be engaged with more than 40 systems vendors in various stages of trials and technical evaluations and has publicly announced a multiyear, multimillion-dollar contract with Innovance Networks.
"There is activity for tunable lasers," muses Arlon Martin, vice president of marketing at Agility. "There are specifications in terms of RFPs [requests for proposal] from tier one companies that require a tunable-laser response. In a market like this, design wins are the most important activity that any of us can be working on."
"It is a make-or-break year for almost any company," concurs Hausken. "Everybody is desperate to get those design wins, because if they don't, how much longer can they go? At least with a design win, you might get something—you can go to your board and say, 'Look, we've been designed-in. Don't give up on us now.'"
One company that definitely hasn't given up on the technology is Intel. In May, the company purchased New Focus's network tunable-laser division for $50 million cash. Why New Focus sold its technology is understandable; the company will focus on tunable lasers for the test and measurement sector and other non-network applications—more lucrative markets in the near term. What isn't so readily apparent is why Intel would enter the market at a time when so many are exiting.
According to Gary Wiseman, director of marketing for Intel's optical platform division, the value proposition was sufficiently compelling. "I can say fairly definitively that we believe that in the not-too-distant future, tunable lasers are going to be the de facto standard for metro and long-haul DWDM systems," he contends.
Wiseman admits that Intel is targeting the early part of 2003 for initial product announcements, and while he would not divulge specific details, he did reveal that New Focus's external-cavity-laser technology met many of the requirements his company was seeking, including high output power, full-band tunability, and "overriding all of these, the confidence that we would have Telcordia-qualifiable, truly production-ready devices," says Wiseman. "We have great confidence that, working with our packaging team, we are going to get to the finish line."
Several tunable-laser manufacturers have taken steps toward Telcordia qualification this year—perhaps a sign that the technology is slowly maturing. Agility met Telcordia's standards in March, followed by Nortel in April. At press time, iolon was targeting September, and both Intel and startup Santur (Fremont, CA) were taking steps to begin the process.
Another indication that the industry hasn't given up on the technology is the work of the Optical Internetworking Forum (OIF—Fremont, CA), which is in the process of developing specifications for common connectors and interfaces for tunable-laser modules. "If history is a guide, that will be an important barrier remover to some implementation issues," reasons Intel's Wiseman.
Others are not so certain, however, that the OIF can remove those barriers alone. "The OIF is a great committee, but it's 40 companies and it's been pretty slow," admits Agility's Martin. "One of our frustrations is that we have to make product, but the OIF doesn't have a standard. We've been working on the side trying to get a subset of some of the major players to agree on an MSA [multisource agreement] that we would then take back to the OIF. But I do think there will be some industry standards during the course of this year."
Look for continued consolidation as well, he adds. At press time, both Nortel and ADC had their optical components divisions up for sale and others may follow suit. There may even be a few new entrants to the market this year. "There are other companies that I'm hearing about, some new ideas out there," observes Hausken. "I don't want to say that the best ideas are yet to come, because I don't know that, but I would say that there are still good new ideas coming up. Which also shows you that the world is not quite ready yet."