Kotura Integrates With $11M
The new direction comes to light as Kotura announces $11 million raised from prior investors ComVentures and Arch Venture Partners. The funding round won't be closed for 60 more days, so the final amount is likely to be higher, says Jean-Louis Malinge, the former Corning executive who was named Kotura's CEO in April (see Kotura Raises $11M, Kotura Names CEO, and Arroyo, Lightcross Join Forces).
"I have refocused everything on integrated optics and silicon," he says. Malinge isn't giving any details of the company's upcoming products, however.
Kotura is shipping a variable optical attenuator (VOA), but its real goal is to pursue integration of optics and electronics, building off Lightcross's work in silicon-based Arrayed Waveguide Gratings (AWGs). The company's material of choice is silicon-on-insulator, which allows it to use conventional complementary metal-oxide semiconductor (CMOS) processes. This kind of manufacturing dominates the semiconductor world and is widely available from foundries such as Taiwan Semiconductor Manufacturing Company (TSMC) (NYSE: TSM).
"The first reaction is 'Cool!' But that and $4.50 will get you a coffee at Starbucks," says Lawrence Gasman, analyst and wag with Communications Industry Researchers Inc. (CIR).
One catch is that CMOS-based optics have proven difficult to make. Previous attempts at this kind of monolithic integration have used polymers or indium phosphide (InP), materials better suited to photonics but not ideal for high-yield, high-volume production.
Another problem is that fab capacity costs money. The "fabless" business model tends to work only when a company is building a high-volume part such as field-programmable gate arrays (FPGAs). As a result, Kotura will have to seek out applications with high-volume potential, Malinge says.
Integration was a hot topic during the bubble, and even afterwards, as a host of companies worked to put active and passive optical components on the same device. Plans such as Kotura's, to integrate electronics, were less common but did crop up from companies including Agere Systems Inc. (NYSE: AGR.A) and Infineon Technologies AG (NYSE/Frankfurt: IFX), Gasman says.
The business case for such "optical chips" was predicated on the rise of all-optical networks, where AWGs would be found all over the network, making them a good platform for integration. But the devices were expensive, and the performance wasn't good enough to justify the price, Gasman says.
Integrated optics might have improved, given time, but startups' money ran out. The fallout has continued into this year: Avanex Corp. (Nasdaq: AVNX) shut down the former Kymata Inc. recently, and Intel Corp. (Nasdaq: INTC) refocused its AWG division to concentrate on enterprise and access markets (see End of the Road for Kymata? and Intel Backs Down on Photonics).
A few companies still carry the integration torch. Earlier this year, Infinera Inc. took the covers off its integrated chip, the heart of its long-haul WDM system (see Infinera Declares WDM War). Tunable-laser firm Agility Communications Inc. also intends to try its luck with integrated InP devices. And Intel is continuing its research into silicon-based lasers, which could provide the foundation for integrated devices.
Integration is "certainly a direction worth pursuing if there can be cost improvements without any degradation, " Gasman says.
Given the difficulty of developing integrated parts, some believe integration is better pursued, not by startups, but by larger companies that can afford to withstand a few R&D cycles without revenues. "It's not clear the first few generations will be taken up by the market," says one venture capitalist requesting anonymity.
So, what could Kotura be building? One obvious approach would be to mash Lightcross's AWG with some electronics, creating a one-chip multiplexer, Gasman says. But he thinks a more promising area would be 10-Gbit/s Ethernet. That's a market crying for price declines, which won't come through volume sales alone; some packaging innovations need to crop up as well, he says. (Gasman adds that he has no idea what Kotura is up to; he just thinks an Ethernet device, if it can be built, could be useful.)
And what about the Arroyo half of Kotura, which concentrated on liquid-crystal tunable filters? Those products have been discontinued. "We have to focus on one particular technology," Malinge says.
Some of the Arroyo employees are still with the company, but Malinge wouldn't say how many, nor would he give a headcount for the entire company.
— Craig Matsumoto, Senior Editor, Light Reading