All-Optical Switch, Take Two

ANAHEIM, Calif. – OFC/NFOEC 2005 – The all-optical switch is dead. Long live the all-optical switch.

Here at the Optical Fiber Communication Conference (OFC), reborn startup Lambda Optical Systems Corp. unveiled its first product, the LambdaNode 2000, an all-optical switch that scales as large as 256x256 ports in one bay.

Yes, all-optical switching is a market fraught with baggage, as service providers have yet to make big, bold moves to adopt the technology. But Lambda thinks it's got a new technological twist that could raise service provider interest: The LambdaNode is the first all-optical switch to combine both switching and transport functions in one system, according to company officials.

Chris Little, Lambda Optical's vice president of hardware engineering, says that LambdaNode's integrated transport capability – as well as advances in the 3D MEMS (micro-electro-mechanical system) technology – will make large all-optical switching work this time around.

"The earlier technology had issues," says Little. "But now there's been a higher level of acceptance of [the] technology by the service providers."

One of those issues in earlier systems, according to Little, was a technical glitch that caused MEMS units to "drift" out of alignment when they became charged with static electricity. Little says that issue has since been fixed. Lambda's core MEMS technology comes from Glimmerglass Networks.

The LambdaNode is designed for metro and regional networks, says Little. The technology pitch of all-optical technology remains the same – by avoiding optical-to-electrical conversion, the switch is more efficient than its OEO (optical-electrical-optical) counterparts, and it can let "express" traffic pass directly through the switch. It is based on a GMPLS control plane and can transport an optical signal as far as 1,000 kilometers without regeneration.

Lambda, a company with 41 employees, isn't exactly a brand-new startup. It rose from the ashes from another startup called FirstWave, which ran into financial problems and ended up in a nasty legal battle with employees before being restructured in 2003 (see Firstwave Given Second Wind).

Of course, Lambda's pitch is likely to be greeted with plenty of skepticism. It's going where several deep-pocketed predecessors – most notably all-optical pioneers Corvis (now a part of Broadwing Corp.) and Calient Networks Inc. – failed before. Those startups heralded a new era of massive mesh optical networks that would require the ability to crossconnect huge numbers of circuits (Calient once touted a switch that could handle 1024x1024 ports). Since then, however, market realities have set in, as carriers never seemed to develop the need for such large crossconnect switching capabilities.

At least one indirect competitor says carriers still don't need large mesh DWDM switches – and are instead focusing on smaller, more flexible add/drop multiplexing technology.

"That's a vision for ten years down the road," said Ashish Vengsarkar, vice president of global development with Mahi Networks Inc. "Most carriers are taking baby steps with ROADM technology."

Mahi sells ROADM gear as well as a metro OEO switch, but doesn't have an integrated all-optical and switching system like the LambdaNode.

But Lambda might be in a position to chip away at the all-optical switching market by landing a few early customers, given its relatively small headcount and the fact that it has fresh funding – $46 million taken in two rounds in 2003 and 2004. Its lead investors include ComVentures and Sevin Rosen Funds (see Lambda Optical Raises $22M).

Lambda Optical says it's already working with at least three potential customers. In addition to a bid for some federal business, reflected in the gear being demonstrated with the Naval Research Laboratory (NRL) here at OFC, Lambda says it's also trialing its gear with two "large incumbent" international carriers.

— R. Scott Raynovich, US Editor, Light Reading

particle_man 12/5/2012 | 3:24:03 AM
re: All-Optical Switch, Take Two The Broadwing NRL application sounds like a long haul or VLH application. Optical switching in that environment creates a pretty constrained and costly network - maybe one reason their margins are under pressure.

If I understand the LO gear, they are aiming at metro and regional applications. You don't have as many challenges (like voas to manage gain tilt, mismatched dispersion profiles, etc.)as you do with LH.
materialgirl 12/5/2012 | 3:24:07 AM
re: All-Optical Switch, Take Two It seems to me that for anyone to want to buy a big optical switch, optical pioneer Broadwing must make a profit. Here is a company with an underused national optical network, and the Navy Research Lab still has to do an OEO conversion to get cross country! Why is that?

Broadwing has 30% gross margins, about the lowest in the transport industry, implying very low pricing given their cost basis. Perhaps optical networks are part of a solution, but not a complete solution. Since optical networks cannot alter the signal in any way, it is my guess Broadwing's prices are very low since they do not "add value." Perhaps a "solution" of some sort needs to be wrapped around this transport.
Sign In