x
Optical/IP

MEMS Put to Test Bed

Ready for another all-optical company?

Subsystems vendor Continuum Photonics Inc. is announcing products today after more than three years of secrecy, but the company isn't banking on an all-optical revival. Its opening move is a switch aimed at the test market.

Based on micro-electro-mechanical system (MEMS), Continuum's DirectLight IG is an optical switch that would sit between test equipment and the devices being tested (line cards, for instance). The switch's programmable nature means engineers could build arbitrary topologies in the lab without rewiring all the equipment being tested. Networks arranged in a ring, for instance, could be converted to a mesh via software, saving time.

Test-equipment vendors are aware of this concept, but it appears Continuum won't face much competition yet. EXFO Electro-Optical Engineering Inc. (Nasdaq: EXFO; Toronto: EXF) was developing its own product based on an all-optical switch, but the project was postponed because the market hadn't yet developed for such a device, officials say.

All-optical switches haven't been used in this space before because they lack the necessary precision, says Aaron Bent, Continuum's vice president of marketing. Test equipment requires switches with near-perfect repeatability, making the market the domain of mechanical steppers sold by DiCon Fiberoptics Inc. and JDS Uniphase Corp. (Nasdaq: JDSU; Toronto: JDU).

Continuum claims it's got the precision and repeatability thanks to its chosen technology. Most MEMS switches use electrostatic or electromagnetic forces to move tiny mirrors, applying a voltage across an air gap. Continuum alters the direction of a beam using a combination of MEMS and piezoelectric forces, "pushing with a material that has the stiffness of aluminum," Bent says. Piezoelectric technology was used by Astarte Fiber Networks, the all-optical switch acquired by Tellium in 2000 (see Tellium's Pure Optical Play).

The piezoelectric method gives Continuum a level of precision usable in test circles, but the company didn't discover this until recently. "We realized about a year ago that we had performance specs that were unique," Bent says.

Another MEMS vendor moving into test applications is Integrated Micromachines Inc. (IMI), one of the startups that preached 1,000-port crossconnects during the bubble. IMI has shifted into the semiconductor probe-card market, the business it had investigated before diving into telecom (see IMMI Switches Strategy).

Unlike IMI, Continuum is still pursuing the telecom market. At Supercomm in June, the company plans to announce DirectLight TL for that market, followed at the end of the year by DirectLight FD, for an application Bent calls fiber distribution.

Systems using the TL probably wouldn't emerge until 2005, Bent says. Along the way, Continuum will have to battle fellow all-optical survivors including Calient Networks Inc., Corvis Corp. (Nasdaq: CORV), Glimmerglass Networks, Lynx Photonic Networks, Movaz Networks Inc., and Polatis Ltd..

DirectLight IG is available in sizes from 16 ports to 64 ports.

— Craig Matsumoto, Senior Editor, Light Reading

Page 1 / 2   >   >>
optical 12/5/2012 | 2:28:43 AM
re: MEMS Put to Test Bed Is 64 ports large enough for the test & measurement market? I would think that large labs would require more ports.

I'd be interested in other opinions.
Shazbot 12/5/2012 | 2:28:35 AM
re: MEMS Put to Test Bed It appears to me that since no telecom company wants to use a 64 port plus optical switch for carrying real traffic, that all the optical switch survivors are putting their eggs in the test market, which is at best small, and will certainly saturate. There are only so many test labs and once they have a switch system they are done.

Glimmerglass is pitching their switch the exact same way. So at least three 64 port vendors have the exact same marketing plan, not exactly "Brilliant". Seems as if all the optical switch vendors that are left are trying to sell a small number of switches to the same people, which means even more will soon die.
fw23 12/5/2012 | 2:28:29 AM
re: MEMS Put to Test Bed I agree. The only true ALL-OPTICAL switch
is built by Corvis.
neomeso 12/5/2012 | 2:28:29 AM
re: MEMS Put to Test Bed This would be a good time to STOP using incorrectly the term "ALL-OPTICAL SWITCH." An all-optical switch has meant for more than twenty years when one light field controls the propagation path of another light field; switching. A MEMS switch is sometimes known as an OOO switch in contrast to an OEO switch, but it is a sloppy use of terms to call an OOO switch an all-optical switch. Now that we are contemplating the use of true all-optical switches for wavelength conversion, there is no room for the misuse of the term. Otherwise, we will soon have the silly sounding case of an all-optical all-optical switch. Hence, Light Reading staff need to show leadership by demonstrating the correct use of terms. I hope not to read any more product brochures or news stories erroneously announcing all-optical switches that are not really all optical, but really optomechanical as in this case. MEMS is still a mechanical device producing the switching, although much more miniature than previous mechanical switches.
optical 12/5/2012 | 2:28:24 AM
re: MEMS Put to Test Bed Corvis the only one to have an ALL-OPTICAL switch....What about Calient?????
fw23 12/5/2012 | 2:28:18 AM
re: MEMS Put to Test Bed Corvis the only one to have an ALL-OPTICAL switch....What about Calient?????


Calient isn't really ALL-OPTICAL. Its a step
along the road to ALL-OPTICAL, but CORVIS
is sixteen steps ahead of it.

Its like comparing the wright brothers initial
airplane with a jet. Callient has the switch
equiveelent of a kite.

Callient is just another Xros.
Junk. Everyone knows it doesn't work. Take
it to the bank.


Lazercat 12/5/2012 | 2:28:14 AM
re: MEMS Put to Test Bed You people need to take a Valium! The terminology "All optical" has been used forever to indicate that the data never undergoes conversion to the electronic switch fabric. How the switch is activated is largely irrelevant to most people. If you want to tout the wonders of your technology start quoting switch time! MEMS are in the millisecond range and "light activated" switches are faster, if they work at all. If you spent more time getting your products to work than you do arguing about what to call them maybe we would see something carry network traffic instead of pushing T&M signals around. Jeeez!
trixie 12/5/2012 | 2:28:12 AM
re: MEMS Put to Test Bed It appears to me that since no telecom company wants to use a 64 port plus optical switch for carrying real traffic, that all the optical switch survivors are putting their eggs in the test market, which is at best small, and will certainly saturate. There are only so many test labs and once they have a switch system they are done.

Glimmerglass is pitching their switch the exact same way. So at least three 64 port vendors have the exact same marketing plan, not exactly "Brilliant". Seems as if all the optical switch vendors that are left are trying to sell a small number of switches to the same people, which means even more will soon die.

----------------------------------------------

Au Contraire-

pay attention at the technical conferences- the architectures are converging- both service providers and equipment vendors are presenting design concepts utilizing hybrid OOO/OEO architectures.

This is a much longer term market, and one that will require a great deal of familiarization and confidence-building with the various systems vendors and customers. One great way to do that is in their labs over the next year or two.

Yes, GG has the same approach. As do others. It will get interesting. The architectures are sufficiently unique that each have their merits.

Do lab managers want to buy huge switch fabrics up front? no. can multiple boxes be hung together and switched in tandem? yes. So modularity and scalability is just as important here as it is in traffic switching. It looks like Continuum, GG and Apcon all offer some form of modularity.

What's the best switch? Depends on what you need it to do.


It will be interesting to see where things land in 6 months.
Raymand 12/5/2012 | 2:27:50 AM
re: MEMS Put to Test Bed I saw the chest beating but I believe misleading given the context ...Only Corvis is ALL OPTICAL..., and the more technicallly minded comments regarding the common misuses of All Optical.

The physical processes that matter from a networking architecture and protocol context are:


Raymand 12/5/2012 | 2:27:49 AM
re: MEMS Put to Test Bed Ooops...Don't use the tab key.

Continuing....

The physical processes that matter from a networking architecture and protocol context are:

1. Switching Speed:
0.01 msec. and longer = circuit switching
Mechanical, MEMS, thermooptic, etceteras.

1 to 100 nanosecond = burst & packet switching
Solid State, SOA, Nonlinear EO, etc.

subnanosecond = bit switching
Solid State, SOA, etc. w/very fast or optical or electronic activations methods.

2. Quantum Carrier Continuity:
Photon-Photon = OOO Avoids OEO conversions
Photon-Electron-Photon = OEO Is OEO converting.
Page 1 / 2   >   >>
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE