Startup Makes Waves in VCSEL Market
The news is significant because, up until now, only two companies claim to have overcome the challenges of manufacturing such lasers.
One is Bandwidth9 Inc., which has a low-cost, low-power (0.2 mW), directly modulated, tunable 1550nm VCSEL targeting metro applications (see Bandwidth9 Cuddles Up With Corning).
The other is Coretek, the startup acquired by Nortel Networks Corp. (NYSE/Toronto: NT) last year for stock then worth $1.43 billion. Coretek has a tunable 1550nm VCSEL that’s pumped with light by another laser, which makes it much more powerful (around 20 mW) – and more expensive. It targets longer-distance applications.
AOI reckons that it can give both Bandwidth9 and Coretek a run for their money, with a high-power and low-cost 1550nm VCSEL, although it declines to say how it'll achieve the feat. “There’s nothing strange involved,” is all that Thompson Lin, AOI's founder and president, will say. It's simply a question of having fine control of the molecular beam epitaxy (MBE) process used to make VCSELs (see Laser Blazers).
All the same, it's going to be some time before AOI starts selling tunable devices.
Right now, the startup is shipping a VCSEL element that goes into an external cavity laser (ECL). The VCSEL element is a "sawn-off" laser. It doesn't need a top mirror because that's provided by the external cavity. As a result of this, the laser can reach higher powers -- 5 mW, says Lin. It's also a lot simpler to make than a complete laser.
This product brought in revenues of $3.2 million last year, he adds.
Lin says AOI's customer for the VCSEL element isn't Blue Sky Research, Iolon Inc., or New Focus Inc. (Nasdaq: NUFO) – which means that there must be yet another highly-secretive startup developing an ECL out there.
Also under development are fixed-wavelength VCSELs -- a product that neither Coretek nor Bandwidth9 considered worthwhile or appropriate for their devices. AOI plans to deliver prototypes of a a fixed-wavelength VCSEL with an output power of 1 mW in October.
A third product, which is also likely to be developed ahead of the standalone tunable, is a VCSEL array. Quite what that would be used for isn't clear.
The big issue with all tunable 1550nm VCSELs is whether they can be produced in large volumes on a production line. Not surprisingly, Bandwidth9 and Cortek say they can do it, but this is questioned by at least one leading researcher in this field, Larry Coldren, a professor at the University of California, Santa Barbara.
Coldren himself claims impressive results for the research his team has carried out on 1550nm VCSELs (see VCSEL Breakthrough at UCal). However, he still says the technology is nowhere near ready for commercial deployment.
The proof of Coldren's convictions are that he helped found Agility Communications Inc., a startup of which he is the CTO (see Agility Packs Three Into One). Agility is developing side-emitting tunable lasers that address some of the same potential markets as those targeted by Bandwidth9, Coretek, and now, possibly, AOI.
— Pauline Rigby, Senior Editor, Light Reading