Calient Networks Inc., one of the few all-optical switch startups left standing after the telecom bust, has unveiled a subsystem version of its product for customized applications.

The product, dubbed the PX, has been in the works for a year and is set to ship in October. Several customers have agreed to buy it, including Montana State University, which is using it in a telecom research lab; and the Photonics Technology Assistance Program, which provides photonic devices for university use and is sponsored by the U.S. National Science Foundation and the Defense Advanced Research Projects Agency.

By offering a subsystem, which can be customized to fit a range of uses, Calient hopes to get other would-be customers "comfortable" with optical switch technology and see how it might work for them. The vendor sees a future for the PX in vendor test labs, as a means of creating a kind of instant optical network, and in grid computing, linking computers and storage. Calient already has its gear installed in a grid project (see Ciena Touts a Piece of the Grid).

The PX is based on the 3D MEMS (micro-electro-mechanical system) technology deployed in Calient's DiamondWave PXC system, an all-optical crossconnect whose announced customers include Chunghwa Telecom Co. Ltd., Japan Telecom Co. Ltd., and KDDI Corp.

At least one analyst thinks Calient's behind schedule in releasing the PX. "Well, it's about time!" says Mark Lutkowitz of consultancy Telecom Pragmatics Inc. "They probably should have done this a year-and-a-half ago. I don't know what held them up."

Indeed, the subsystem route is well worn by other all-optical switch startups. It's led to mixed results: Some have disappeared, like the late OMM Inc. (see OMM: The End Is Near) and Network Photonics (see Network Photonics Shuts Down). Altamar, another company that sold its wares as optical subsystems, was bought by JDS Uniphase Corp. (Nasdaq: JDSU; Toronto: JDU) in July (see JDS Uniphase Picks Up Altamar).

Calient CEO Charles Corbalis says the company hadn't been convinced that sustaining a separate subsystem business would be worth it. Now, though, he thinks the prime prospects for the PX, namely, telecom equipment makers, are kicking the tires on optical switching again, after putting the technology on hold for a couple of years.

Calient is different from its failed predecessors in several important ways, Corbalis says. "We were fortunate to get a product to work," he says wryly. Also, Calient's technology supports up to 256x256 ports, making it larger than others that have tried this market and failed, he says.

Finally, Calient isn't giving up on its system business in order to move to a subsystem product, Corbalis says.

Still, there are challenges ahead. Based on the preliminary customer announcements, it doesn't exactly sound like a blockbuster market. Real growth may take a long time. What's more, the prospects for subsystem use may take a while to solidify. After all, grid computing is still a relatively new concept, even though prospects are good (see Grid Networking).

"There's no way to tell when the market will take off. But this at least reduces risk and puts [Calient] in a more healthy position -- attractive to a system vendor who may want to buy them down the road," Lutkowitz says.

— Mary Jander, Senior Editor, Light Reading