An Optical Race with No Winners -- For Now

Nortel, Lucent, and Alcatel are building 40-gigabit networking systems that offer more capacity than most customers currently clamor for

When Nortel Networks (NT ) came out with the fastest ever optical system in 1996, the rollout elicited nothing more than a shrug from executives at chief competitor Lucent Technologies (LU ). Less-advanced systems weren't even close to capacity yet, they figured.

But much of Nortel's stunning success over the past few years springs from that move. The company currently holds the lion's share of the market -- 90% -- for those fast so-called OC 192 systems, which allow data-transfer speeds of 10 gigabits per second. The battle for the 10-gigabit market is over -- and Nortel has won. But another clash, over OC 192's more powerful successor, is unfolding. And it's starting to look like lightning might not strike twice for Nortel.

It's already feeling the effects of the telecom spending pullback. Nortel announced, after the market closed on Mar. 27, that its first-quarter earnings and revenue numbers would be much lower than expected. It's the company's second warning in two months. Nortel expects to report a loss of from 10 cents to 12 cents a share, on revenue of $6.1 billion to $6.2 billion. That comes after its anticipation on Feb. 15 of an operating loss of 4 cents a share on revenues of $6.3 billion. The warning is likely to put further pressure on the stock, which on Mar. 27 traded at around $17, down from $89 in July.

Seems like bad timing, because on Mar. 19, Nortel unveiled its first 40-gigabit system (offering four times the capacity of the 10-gigabit). "We've got the track record," insists Nortel Optical Networks President Greg Mumford. The company is due to make its 40-gig systems commercially available in the first quarter of 2002. But other optical players are expected to do the same. They aren't about to let Nortel get ahead again, explains analyst Karen Liu, analyst at RHK Inc.


  Still, rival equipment makers, which are rushing to block Nortel from gaining the first mover's advantage, may be in the right place -- at the wrong time. These companies have spent billions creating and building systems that carriers won't be able to afford until the economy and the financial markets recover, analysts say.

Last November, Lucent began 40-gig trials with voice, data, and Internet services provider Global Crossing (GX ). Now, Lucent is conducting a trial in Europe. "Everything looks very promising in terms of stability of the systems and major indicators," says Olaf Herr, product manager of Lucent's optical-networking group. "We'll make sure that [the 10-gigabit-system fiasco] won't happen again." Alcatel, which conducted its first trial with Dutch telco KPN Quest in the third quarter of 2000, has already completed all trials, says Karl Traberg, director of marketing for optical networking at Alcatel (ALA ).

But this is about far more than a competitive race to market. Courtesy of a battered stock market and slowing economy, this could be an area with no winners for several years -- and one that could leave optical-gear companies, desperate for new moneymaking products, holding the bag for massive research and development costs with no immediate return. "We see the migration from 10 gigabits to 40 gigabits as slow migration, not as a radical change," says ABN AMRO analyst Ken Leon. That's a troubling outlook for this downtrodden sector.


  Existing systems offer telecoms enough capacity to accommodate current traffic, says Tom Lauria, analyst with ING Baring. Facing tight capital markets, heavy debt loads, and pared-down plans for infrastructure improvements, carriers likely won't shell out big bucks for the new technology before they have to -- possibly three or four years from now, Lauria says. That's in sharp contrast to '96, when capital was flowing like water, telecoms could do no wrong, and investors were willing to finance technologies based on the faith that if they build it, customers will come.

Now, carriers needing extra capacity might opt for an additional 10-gigabit system instead. Increased competition has brought prices for 10-gig systems down significantly, says Lauria. At the same time, before the new, 40-gig systems see wide deployment, technological barriers have to be overcome as well.

Older optical fiber already in the ground might not be able to handle 40-gig transmission, says Shin Umeda, principal analyst at networking consultancy Dell'Oro Group. And 40-gig signals would have to be reamplified every 124 to 186 miles -- far more often than carriers would prefer, says Traberg.

To be sure, all of these problems -- driven by technical, financial, and market-demand issues -- will be worked out in a few years. That's when telecoms will start buying. And 40-gig systems promise significant savings in data-transfer costs when their prices come down.


  Investors betting long term on gearmakers Nortel, Alcatel, or Lucent likely won't be disappointed. Sooner or later, the technology will catch on. In fact, 40-gig is projected to capture 20% of the $57 billion optical-transport market by 2005, according to Dell'Oro.

As high-bandwidth entertainment and remote teleconferencing become commonplace, demand for these ultra-high-speed systems will materialize -- in as soon as three years. At that time, optical systems with more capacity will be needed to move that heavy info load around with relative ease, says Bill Lennon, an engineer at Lawrence Livermore National Laboratory. The need for the technology "is going to sneak up on us," he says.

However, in an environment obsessed with profits and revenue growth, three to five years is a lifetime. In the meantime, the companies' R&D costs aren't likely to generate much return -- something that probably will further sour the spirits of many Wall Street tech investors.

By Olga Kharif in New York

Edited by Beth Belton

Before it's here, it's on the Bloomberg Terminal.