PIC startup makes good on $19.5 million promise

MARCH 17, 2009 By Stephen Hardy -- Startup OneChip Photonics has earned a second and final tranche of VC funding by convincing its investors that it can commercially produce transceiver PICs.

MARCH 17, 2009 By Stephen Hardy -- Startup OneChip Photonics, which is developing InP-based photonic integrated circuits (PICs) that monolithically integrate all of the functions of a transceiver in a single device, managed to arrange $19.5 million in VC funding back at the end of 2007. However, the funding came with a catch: Half would be available immediately, to enable the company to develop its technology. But they wouldn't receive the other half until they could prove to the investors' satisfaction that the technology was commercially producible.

OneChip announced today that they successfully achieved that milestone. Now all they have to do is prove the same thing to potential customers.

Jim Hjartarson, CEO of OneChip Photonics, likens his company's technology to that of Infinera (search Lightwave for Infinera). Yet while Infinera has ridden its InP-based PICs to a significant share of the metro and long-haul markets, Hjartarson says he plans to focus on very high volume, low cost applications. Initially that will mean FTTH; the company plans to have prototype transceivers for both EPON and GPON architectures within the next three to six months.

OneChip's potential cost advantages lie in two areas, Hjartarson says. First is in the fact that the company's PIC technology requires only a single epitaxial growth step; the creation of the rest of the transceiver components and functions are just a matter of etching and photolithography, he says. These components and functions include all the active and passive elements required for a transceiver, he asserts: DFB laser, optically pre-amplified detector (OPAD), wavelength splitter, spot-size converter, and various passives.

Hjartarson declined to reveal the size of the wafers the company uses, except to say that they're "a fair bit larger" than the 2-inch and 3-inch wafers commonly associated with InP. The company uses two outside partners to create the wafers whose identities he likewise would not reveal.

The second cost reduction area derives from the fact that, since the transceiver components reside in a single integrated device, the optical assembly processes normally associated with transceivers are no longer necessary. The only function currently not completely automated is the attachment of the fiber to the chip, and Hjartarson says he expects to have that process automated by the time the company's transceivers reach full production.

While the company touts the costs savings their approach provides versus conventional transceivers, Hjartarson wouldn't quantify that amount. He did say that the cost differences would be particularly pronounced at speeds of 2.5 Gbps and higher, and acknowledged that devices for these speeds are on the roadmap. He asserts that one potential customer said OneChip's technology would be the difference between making money on ONTs or losing it.

In particular, Hjartarson says the integrated OPAD technology offers significant cost advantages over the use of separate avalanche photodiodes.

Hjartarson says that the latest investment tranche -- which came from BDC Venture Capital, DCM, GrowthWorks Canadian Fund, and Morgenthaler Ventures -- should meet the company's needs for the foreseeable future, at least as far as FTTH is concerned. However, Hjartarson says he has his eyes on other opportunities, including consumer and data applications, and a quick move into those markets might require additional funding.

Visit OneChip Photonics

More in Transmission