& cplSiteName &

Onix: Another MEMS Casualty

Light Reading
News Analysis
Light Reading
5/22/2003

With the bang of a gavel earlier this week, the book closed on Onix Microsystems Inc. (not to be confused with the former chip startup Onex -- see Onex Chip Sees Light of Day), the latest casualty in the formerly high-profile business of all-optical switches.

The last of Onix's assets went to auction Tuesday. Whether the company's intellectual property found a home is unknown, as the organizers haven't yet disclosed all the results to Gary Koos, Onix's vice president of finance and its last remaining employee.

In its prime, Onix employed as many as 250. "We went down to 20 back in October and 10 in January. Now we're down to one," Koos says. He confirmed to Light Reading that the company has shut down.

In 2000, photonic-switch companies were all the rage. Well more than a dozen companies began developing all-optical switch fabrics, usually based on MEMS (micro-electro-mechanical system) or liquid-crystal technology but sometimes dipping into more exotic technologies (see Optical Switching Fabric).

Like OMM Inc., Onix raised a bundle of cash thanks to the hype (see MEMS Startup Onix Gets $35.5M and Onix Scores $95 Million). And, like OMM, Onix had to scale back its ambitions once it became clear that large all-optical switches weren't a viable market (see Onix Follows in OMM's Footsteps). And now, Onix is following OMM into oblivion (see OMM: The End Is Near).

Tiny all-optical switches, such as 1x2 devices for protection switching, continue to be a viable market, with players such as JDS Uniphase Corp. (Nasdaq: JDSU; Toronto: JDU). But Onix and others were chasing bigger game, the large-scale all-optical switch that would someday power all-optical networks.

Network Photonics was another recent casualty there, having offered its technology for sale last month (see Network Photonics Shuts Down). And other MEMS hopefuls such as K2 Optronics Inc. and Nayna Networks Inc. stepped away from the all-optical market some time ago, although both companies are still in business.

Among the survivors nowadays is Nanogear Inc., a company created in December when its founders acquired MEMS technology from the defunct C Speed Corp. Nanogear appears to still be in stealth mode and did not immediately return a call for comment.

Other survivors such as Lynx Photonic Networks have scaled down their ambitions. The company cut back its work force in October but still has 56 employees. "We feel that's sustainable," says Michael Leigh, president.

Lynx, which sells pizza-box subsystems, doesn't make Optical Crossconnects larger than 8x8. It can produce a 16x16 version, but the demand isn't there -- for the few requests requiring anything larger, Lynx finds it's sufficient to chain together multiple small fabrics. "We've been able to do a lot of scaleability with our small matrices," Leigh says.

Advanced Optical MEMS Inc. (AOMEMS) likewise continues to sell 4x4 and 8x8 optical crossconnects. As with Lynx, one of its major markets is protection switching.

But don't fret -- there are still a few hearties going after large-scale all-optical routing. Calient Networks Inc. and Movaz Networks Inc. have not only survived, but have done so using difficult 3D MEMS technology (for a complete description, see Optical Switch Fabric -- MEMS).

Calient has said its DiamondWave switches are shipping with optical crossconnects ranging from 32x32 to 256x256. The company plans to make some customer announcements in conjunction with next month's Supercomm tradeshow in Atlanta.

— Craig Matsumoto, Senior Editor, Light Reading For extensive and up-to-date coverage of Supercomm – before, during, and after the show – visit Light Reading's Supercomm Preview Site.

(25)  | 
Comment  | 
Print  | 
Newest First  |  Oldest First  |  Threaded View        ADD A COMMENT
Page 1 / 3   >   >>
single mode figure
single mode figure
12/5/2012 | 12:01:09 AM
re: Onix: Another MEMS Casualty
I doubt historians in tech will forget this period as a mem's bubble, from the component people to expensive experiments like XROS, some people made money and had a great ride. But most of us are formerly almost rich class. Too bad life is unfair....
on the cusp
on the cusp
12/5/2012 | 12:01:08 AM
re: Onix: Another MEMS Casualty
MEMS is good for small switches as LR article says, but cannot be made to work or has no business case for large applications. Looks like near term winner is liquid crystal. Not a lot of hype but a number of companies making solid progress both for small devices and multiwavelength applications.
austin_powers
austin_powers
12/5/2012 | 12:01:08 AM
re: Onix: Another MEMS Casualty

I looked at Onix as a possible supplier. In their whole history, they never seem to be able to execute though it seems. Unlike OMM (on component level) and Calient (on systems level), Onix was never able to produce a sample or prototype of anything they had on their product roadmap (either 2D MEMS or 3D MEMS technology). Their investors must have been very frustrated because Onix over many years of being around never brought anything out to the marketplace even though they promised customers a lot. Not sure if it was management, the people they hired, or what, but this was a clear case of poor execution on all fronts. At least OMM, Calient and even Lucent (with the lambda router) showed "something" for the money that was invested.
zettabit
zettabit
12/5/2012 | 12:01:06 AM
re: Onix: Another MEMS Casualty
My first inclination would be to expect Calient's customer announcement to be:

"we have none"

However that's probably not what they were intending.

$10 says they announce some type of government contract.

$5 says they announce a trial with NTT

However my money is on them being gone in N months.

N = $ in bank / monthly burn rate
deer_in_the_light
deer_in_the_light
12/5/2012 | 12:01:05 AM
re: Onix: Another MEMS Casualty
Might be an interesting play for the MOEMS survivor.

Glimmerglass has interesting 3D and is well positioned on the high-end after the deaths of OMM, Transparent, Onix, Network Photonics.

Lightconnect is covering the mid-range WSS, blocker and DGFF market. JDSU is gradually getting there and is moving away from liquid crystal.

On the low-end Dicon owns most of the 1x2 2x2 switch market, I am not sure what is happening on the MEMS VOA front, there are a lot of established companies there like Dicon, JDSU, Lightconnect but I keep seeing new guys trying to break in (Santec, MEMSCap, Chromux). I guess that's what selling today and everybody is trying to get a piece of the action.

Still, consolidation is badly needed in MOEMS.





Shazbot
Shazbot
12/5/2012 | 12:01:04 AM
re: Onix: Another MEMS Casualty
And you work for a....let me guess... could it be an LCD switch startup? Maybe in Marketing? LCDs are inherently polarization affecting devices, wouldn't they have PMD and PML problems in most cases. MEMS certainly can be made to work for large cases, but I agree there is no present business case for them.
rapid
rapid
12/5/2012 | 12:01:03 AM
re: Onix: Another MEMS Casualty
Dear Shazbot,

He or she doesn't, but I do...I suggest that you get your technical knowledge up-to-date regarding LC technology...polarization affecting? maybe for neophytes!

I bet you still think that LC materials are slow, undependable, and temperature dependent?

Lack of technical awareness is second only to lack of vision as a hi-tek sin!

y2k
y2k
12/5/2012 | 12:01:02 AM
re: Onix: Another MEMS Casualty
Austin_Powers: I agree with your assessment of Onix. I know of someone who had a chance to take a close look from the inside. My understanding is that their engineers are well intentioned people. Unfortunately, they were mostly fresh Ph.D.Gs with practically no product experience.

On the other hand, the senior management had product experience but no startup experience and really had no idea how to execute unless it was done at least once. So there were disconnects. The people were not necessarily incompetent, just miscast.

Their investors should share a big part of the blame as well. They fired their first CEO only after it was too late but their second CEO was even more pathetic. He had no relevant experience in telecom, had never ran a startup in his life and by the way, was boot out of his previous company because he took $450M cash and ran it bankruptcy.

The Silicon Valley is full of engineers and entrepreneurs who have failed at least once. But for the most part, especially during the last two years, we could always point to technical success and blame the rest on the market.

Onix failed not because of the market, they failed because they could not execute and in the end, they had to return over $25M to their investors because no one could be convinced anymore that they know what to do with them.

Very few of us have benefited financially during the bubble, but at least some of us benefited professionally. This is probably true if you worked for OMM or Calient, unfortunately much less so if you worked for Onix. It is too bad,
Demander
Demander
12/5/2012 | 12:01:01 AM
re: Onix: Another MEMS Casualty
Rapid says,
"I bet you still think that LC materials are slow, undependable, and
temperature dependent?"

The interesting thing is, when the MEMS products acutally came out it seemed they weren't any faster than LCs!

So just how fast, dependable, and temperature insensitive are they now? Are the speeds quoted only the voltage "on" speeds? Be honest now :-)

Is the birefringe less temperature sensitive than it ever was or do the designs compensate for it?

You don't have to convince me about dependability. I'd take LC over mems any day. I've had too many displays outlast their calculators, watches, etc.

Demander
redface
redface
12/5/2012 | 12:01:00 AM
re: Onix: Another MEMS Casualty
"Onix failed not because of the market, they failed because they could not execute and in the end, they had to return over $25M to their investors because no one could be convinced anymore that they know what to do with them."

Well said.

I recently had a chance to inspect the Onix switch under microscope. While it is tiny and very cute, it seems to have a fatal flaw: the mirror arms are very easy to break because they are based on flexure mechanism and they need to be twisted by 90 degree over a very short distance. The switch I inspected had almost 90% of mirrors missing, which I suspect was due to this problem. So maybe it was technical difficulty which doomed Onix.

Could any insider confirm this? And, I thought Onix switch is based on electrostatic actuation, but I am perplexed by what seems to be an electromagnet (or a heater) inside the switch package. What's going on there?
Page 1 / 3   >   >>
Featured Video
Upcoming Live Events
November 5, 2019, London, England
November 7, 2019, London, UK
November 14, 2019, Maritim Hotel, Berlin
December 3-5, 2019, Vienna, Austria
December 3, 2019, New York, New York
March 16-18, 2020, Embassy Suites, Denver, Colorado
May 18-20, 2020, Irving Convention Center, Dallas, TX
All Upcoming Live Events