CMOS Grows on Clarendon

Plenty of companies developing integrated optical components have, at one time or another, claimed they were going to recreate the runaway success of the electronics industry. (You know, the one that was revolutionized by mass production of integrated circuits?)

In most cases, of course, they've been stretching the truth a little too far (twang!).

However, Clarendon Photonics Inc., a startup that emerged from stealth mode today, appears to be taking the parallels with electronics further than most (see Clarendon, Micron Announce Multiplexer).

For their kickoff, Clarendon is announcing a partnership with a foundry, Micron Technology Inc., to manufacture its products. It's pretty normal for integrated-optics firms to make components using semiconductor process equipment. And it's not unusual for them to outsource manufacturing: A number of specialist optics foundries exist, including Denmark's Ionas A/S and the U.K.'s Optical Micro Devices Inc. (OMD). But it is unusual for integrated optics to be manufactured on a standard CMOS production line, as Clarendon's are.

The use of existing CMOS (complementary metal-oxide semiconductor) production lines is a key part of Clarendon's business strategy, says Pierre Villeneuve, the startup's founder and CEO. "It allows us to build a company without having to invest in expensive capital equipment," he explains.

Villeneuve says that the partnership with Micron, which was established in August 2001 to develop CMOS-compatible fabrication processes for integrated photonics, has now come to fruition. The first prototype off the production line is a four-channel reconfigurable optical add/drop multiplexer (OADM), which will be showcased at the upcoming Optical Fiber Communication Conference and Exhibit (OFC).

The overall reconfigurable OADM solution occupies a 3x4-inch data board, which is considerably smaller than existing OADMs, according to Bill Thompson, Clarendon's VP of marketing. "We've seen as many as three cards required to build a four-channel OADM," he contends.

Competition for this type of component is coming from several directions at once. The incumbent seems to be Corning Inc. (NYSE: GLW), which makes a reconfigurable OADM based on liquid crystal technology (see Optical Switching Gets Flexible). Marconi PLC (Nasdaq/London: MONI) uses Corning's components inside its reconfigurable OADM system, the PMA-32. In addition, Agilent Technologies Inc. (NYSE: A) has its "bubble switch," which is being deployed by Alcatel SA (NYSE: ALA; Paris: CGEP:PA). Further competition is emerging from vendors building components based on Arrayed Waveguide Gratings (AWGs) and optical switches, such as Lynx Photonic Networks (see Look Ma, No Moving Parts!).

Another way in which Clarendon is making the manufacture of optical components more like that of electronics lies in the architecture of its devices. Just as digital ICs are built by taking individual building blocks -- transistors -- and making connections between them to define the application, Clarendon plans to make a range of products based on a unique building block that can be arranged in different configurations to perform different functions. Other integrated-optics firms require different designs for each function.

Clarendon calls its building block a "DMR optical processor", where DMR stands for dynamic, multifunctional, reconfigurable (no wonder they abbreviated it). Measuring 200 microns long and a few tens of microns in diameter, it is based on buried waveguides, waveguide coupling, and optical gratings, which can be thermally tuned to operate at different wavelengths. Each DMR optical processor is physically the same, except for the design and control of the gratings, the company claims. By changing the properties of the grating, it's possible to implement different functions, such as variable optical attenuators and couplers.

The idea behind the DMR unit was born at the Massachusetts Institute of Technology (MIT), where Villeneuve was working on photonic crystals: nanostructured devices that offer new ways of controlling light (see The Hole Thing). "I realized that we didn't need photonic crystals to make frequency-selective circuits," he says. "We can do the same things with existing components like gratings."

Villeneuve claims Clarendon holds exclusive rights to some of the patents dreamt up by the photonic-crystal researchers at MIT, and the patents are applicable to a wider range of components than just photonic crystals. "They were written in a pretty broad manner. I should know -- I wrote them," he chortles in his joy.

Clarendon has raised $20 million to date from Ampersand Ventures, New Enterprise Associates (NEA), Sevin Rosen Funds, and Vortex Partners. It intends to start raising a series C round later this year, after customer trials of its OADMs are completed. The company projects trials should start in four to five weeks.

— Pauline Rigby, Senior Editor, Light Reading
http://www.lightreading.com For more information on OFC 2002, please visit: www.nottheofc.com

Page 1 / 3   >   >>
BB 12/4/2012 | 10:51:04 PM
re: CMOS Grows on Clarendon i interviewed with these chumps a while back....boy are they abunch of chumps or what...they had no idea what they were doing...took the ceo, cto and others about half an hour to get the projector going.... basically a bunch of university lab rats... who think they're all that just cause they got money thrown at them during the heyday....they're not worth the time that it took me to write this message...

my honest $0.02
mariner 12/4/2012 | 10:51:01 PM
re: CMOS Grows on Clarendon Sounds like you didn't get the offer BB.
mbock 12/4/2012 | 10:51:00 PM
re: CMOS Grows on Clarendon Well I interviewed with them too and I thought they were very good. However, I decided to stay where I was. I think they may surprise just about everyone.
lightfree 12/4/2012 | 10:51:00 PM
re: CMOS Grows on Clarendon If you think they were very good, and that they may surprise everyone (read: they may be sucessful), then why didn't you join???
BB 12/4/2012 | 10:50:59 PM
re: CMOS Grows on Clarendon actually i did get the offer, but decided to go elsewhere...they were just too "graceful" for me.
[email protected] 12/4/2012 | 10:50:54 PM
re: CMOS Grows on Clarendon Hey BB and Mbock

Having had the (doubtful) pleasure of interacting with the delightful P.V when they started out I feel that they obviously have great academic credentials etc but question whether or not a very smart Researcher can create an atmosphere that is condusive to delivering great commerical products.

Any thoughts

BB 12/4/2012 | 10:50:53 PM
re: CMOS Grows on Clarendon dwdm guy,

yeah couldn't agree with you more... P.V is an embarassment...he may have a future in physical comedy. like i said before....they saw everyone else getting rich and decided to go for it too...if i had met these guys before the bubble burst i would have forcasted the bubble burst...

PV is a dork with a capital D.
Pauline Rigby 12/4/2012 | 10:50:39 PM
re: CMOS Grows on Clarendon Clarendon is using silica-on-silicon like just about everyone else. The silicon is just there for structural support, and the waveguide is made on top of that. CMOS processing also requires the deposition of silica, to insulate under the transistors (I think that's right -- someone correct me if it's not.)

[email protected]
willsonwr 12/4/2012 | 10:50:39 PM
re: CMOS Grows on Clarendon On a lighter note - what material system are they supposedly using? Presumably silicon if using Micron.

mbock 12/4/2012 | 10:50:35 PM
re: CMOS Grows on Clarendon Consider this...what if P.V. recognized his limitations and acquired the right team to deliver the goods.

It takes a smart researcher to make a prototype that works but it takes a an even smarter researcher to know how to hand it off and who to hand it off too to create the commercial product.

My impression of P.V. is that he had a good grasp of that concept.
Page 1 / 3   >   >>
Sign In