x
Optical/IP

Movaz Sees 3D

If they hand out moxie awards at next week's Optical Fiber Communication Conference and Exhibit (OFC), Movaz Networks Inc. should at least be in the running.

The latest from this daredevil startup: At the show, it's planning to demo an alpha version of the module it will use in its optical crossconnect product, due later this year -- a module that's based on so-called 3D MEMS (micro-electro-mechanical systems).

Movaz announced the module, dubbed the integrated Wavelength Selective Switch (iWSS), today (see Movaz Debuts All-Optical Switch). It intends to show a 400x400 version at OFC, made with technology it owns through the ten-year agreement it signed last year with The BF Goodrich Company (NYSE: GR) (see Movaz on the Move).

This takes real cojones. For one thing, carriers right now are about as interested in optical switching as they are in gaining third nostrils. For another, 3D is risky business these days.

"3D is still expensive and unproven," asserts Conrad Burke, senior VP of sales and marketing at OMM Inc., a maker of MEMS components.

At issue is the MEMS technique of using tiny tilting mirrors to guide light through a switch. In 3D MEMS, at least until recently, the approach involved setting up rows of mirrors to move in multiple planes, requiring several subparts that are difficult to design and manufacture efficiently.

Worse, early 3D MEMS switches reportedly suffer from very bad crosstalk -- signal interference that results as the mirrors move from one position to the next. This is reportedly what pulled down development of the Xros switch at Nortel Networks Corp. (NYSE/Toronto: NT), an effort that was finally ditched recently (see Nortel Shuts Optical Switch Effort).

For these reasons, 3D MEMS is seen by many as a shaky proposition. Several components companies, including OMM and startup Umachines Inc., say so-called 2D MEMS, wherein mirrors are moving along only one plane, is superior: It's faster, simpler, easier to manufacture, and still perfectly suited to the needs of carriers over the next couple of years.

Movaz thinks differently. By simplifying the design, it hopes to avoid the difficulties Nortel faced with Xros. "Xros deployed three different devices. We must use one module to demux wavelengths from fiber, switch them, and mux them back into the fiber," says Bijan Khosravi, CEO of Movaz. Furthermore, he insists that the proprietary technique developed by Movaz and Goodrich, one which uses freespace optics to mux and demux wavelengths, eliminates crosstalk and other negative effects suffered by 3D MEMS of the past.

He's also confident his MEMS will be easy to make. "We knew the process to build was as important as the design itself," explains Khosravi. For this reason, he says, Movaz has undergone nine separate cycles of design, prototype, and manufacture with BF Goodrich to ensure the new switch will be producible.

Nor is Khosravi worried about relying so heavily on just one MEMS manufacturer. "Goodrich is so entrenched," he says. "They're a huge company."

The small circle of 3D MEMS suppliers may be growing: Earlier this week, Agere Systems (NYSE: AGR) unveiled 3D MEMS components (see Agere Ramps Up 3D MEMS). Transparent Networks Inc. just announced a very large 3D MEMS switch. And Integrated Micromachines Inc. (IMMI) is planning to unveil an 80x80-wavelength switch module based on 3D MEMS at OFC next week. As is the case with Movaz, it remains to be seen how well these emerging solutions overcome the obstacles that have hindered 3D MEMS up to now.

How will the new switch fit into the rest of Movaz's wares? As a high-end augmentation, says Khosravi. While the RAYstar, a 320-wavelength, electrically-based central office switch announced last month, is the company's flagship product, Movaz will fill specific requests for core networking with an all-optical switch based on the iWSS, which is set to debut later this year (see Movaz Makes Its Milestones).

"Our solution offers transport, switching, and control plane features," Khosravi says. He claims he's already got an international carrier and two RBOCs "anxiously waiting" to see the prototypes.

Movaz could use a high-profile carrier contract. While the startup has hit all delivery dates and milestones so far and closed a range of deals with small independent carriers worldwide, it's failed to produce a contract with a key player. What's more, an early defection by Genuity Inc. (Nasdaq: GENU) as a trial customer (reasons undisclosed) didn't help matters.

But, as ever, Khosravi remains undaunted. "We're moving out of the development phase we've been in for 18 months," he crows. "Now we want to grow our business and focus on the RBOCs."

The company has taken a major step in this direction: Within the next few weeks, it's set to announce its Telcordia Osmine certification, a key requisite for any RBOC deal.

Movaz also has a new president and COO, Guy Gill, cofounder and former CEO of Elastic Networks Inc. (Nasdaq: ELAS), who also had a long career at Nortel.

Gill is as enthused as Khosravi: "We have a great unified platform, with the economics to allow service providers to reduce their costs," he says.

Nerves of steel help, too.

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

Page 1 / 3   >   >>
self 12/4/2012 | 10:37:15 PM
re: Movaz Sees 3D Well, here is one that just popped up.

But, I have a hard time seeing what is special about this except it is a nicely, tightly integrated system. The MEMS are standard double gimbals. The only difference is that they are better at rotating in one direction than the other. But, they are probably so closely tuned to this specific implementation that they have no use beyond this Movaz box.

WO 02/25358 A2

<http: bnsviewer?cy="gb&amp;LG=en&amp;DB=EPD&amp;PN=WO0225358&amp;ID=WO+++0225358A2+I+" dips="" l2.espacenet.com="">
</http:>
self 12/4/2012 | 10:47:33 PM
re: Movaz Sees 3D To see whether they (or their license holders) have made they same mistakes as most people.
closends 12/4/2012 | 10:47:34 PM
re: Movaz Sees 3D Talked with friend who knows some of this story. It really is 3-D. Says story is correct as written. but hyped a little. Patent is a license deal. Couple of real breakthroughs involved. Packaging described as "embarassingly simple".. Optics also simple, cheap and apparently quite small. Big team of people did it; they have 300 folks! Including some folks I had lost track of. He confirms Movaz is very serious about GMPLS, and this device fits their plans.

I searched LightReading and found this from a month ago:

Article Talk

Movaz Makes Its Milestones
Author: y2kNumber: 11
Subject: Re: MEMS fabricDate: 2/8/2002 1:15:44 PM


I was in the audience when their CEO gave a presentation at an Analyst meeting. He talked about their 3D-MEMS Gǣwavelength-selectiveGǥ switch and their partnership with Harris and Goodrich. There was a top-view picture of the NxN device (I donGt remember what N is, but I think he said that it is at least 400).
Harris has been building CWDM system for military aerospace application using free-space diffracting grating and the CEO said that they are partnering with them to incorporate the mux/demux technology into their switch. In addition, they are partnering with Goodrich (which had acquired a company called Advanced Micromachines) to make the MEMS mirrors.
So in theory, this box will take one or two input fibers, completely demux the wavelengths, switch them at will, drop and add wavelength-level traffic at will, re-mux and send them out on one or two fibers. The CEO was very confident and said they will demonstrate working system at OFC.
Someone asked him about insertion loss but I donGt remember if he answered or not. Afterwards, I overheard someone said that it is at least 8 dB but I have no idea if this is true or not.
This will be an amazing accomplishment and I think it is worth watching. Too bad that I donGt have any discretional money left, otherwise I will put it all on this one because they apparently have single-handedly solved the entire suites of combined problems that at least two dozen well-funded startups couldnGt solve individually.
--y2k-

That guy sounds like their marketing department. But could they live with 8 db? We thought that was at the limit.

Thanks for feedback on Tellium.

Anyone going to be at OFC to give some untainted feedback on reception??
alloeo 12/4/2012 | 10:47:35 PM
re: Movaz Sees 3D It could be from anyone... universities, private parties, etc.

Of course you won't find much under the companies' names, unless it's a large established company that performs R&D. Startups like these rarely have patents under the their own name. The patents should have been filed way before the company comes into existence.

If you search the literatures, the MEMS clan is a pretty small community. You can also do the USPTO search.

After all, why do you want to dig out whose patents they are using?
self 12/4/2012 | 10:47:39 PM
re: Movaz Sees 3D they license patents from others.
____________________________________

But from whom? I can't find anything under BF Goodrich, Advanced Micromachines (owned by Goodrich), or Movaz.
befuddled 12/4/2012 | 10:47:53 PM
re: Movaz Sees 3D Say an announcement this morning touting Guy Gill joining on at Movaz. Sounds like a reunion of the Nortel gang. Wonder how long Guy will like reporting to Bijan? Who's next to join Movaz, Anil Khatod?
alloeo 12/4/2012 | 10:48:00 PM
re: Movaz Sees 3D they license patents from others.
hyperunner 12/4/2012 | 10:48:00 PM
re: Movaz Sees 3D ...and isn't Tellium already doing switching with mirrors?

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

No, Tellium's Aurora Optical Switch is an OEO switch that has OC-48 grooming granularity.

Like several other companies, they have recently announced that they'll build a hybrid OOO/OEO architecture based on 2D MEMS. A while back Tellium acquired a 3D MEMS company (Astarte). I guess it was an expensive way to find out that 2D MEMS is a more realistic option :-)

hR
closends 12/4/2012 | 10:48:01 PM
re: Movaz Sees 3D The Nortel/Xros story was painful, even if you didn't own stock. Hell of a costly mess. But I still think a MEMS solution has cost going for it: it IS the next generation...and isn't Tellium already doing switching with mirrors?

The problem with this story is that the Movaz PR department has surely overreached. Specsmanship lives? Their web site is telling us zip; they need to tell us more before we should believe. But,if they can do what they say, who cares if it is 1-D, 2-D, 3-D or 4-D? BF Goodrich has an optical outfit that makes spy cameras and such. Maybe they figured this thing out? Taxpayer dollars at work?
self 12/4/2012 | 10:48:06 PM
re: Movaz Sees 3D The port counts sounded similar. Network Photonics was saying 8 ports of DWDM in - Drop Mux - Lots of Switches (?) - Add Mux - 8 ports out. (Corvis like?). Is the Movaz architecture similar?

_______________________________

NP likes to play games confusing the number of wavelengths they can pass through their switch fabric with the number of places to which they can switch them.

After they demux a bundle of wavelengths (say 40) each wavelength hits one specific 2D mirror on a 1x40 row of 2D mirrors. Each mirror can point at one of only two ports (eg Add or Drop).

This is very different from the designs of traditional OOO switch fabric makers, either 2D (ADC, Advanced Optical MEMS, Dicon, JDSU, Northrop Litton Poly-Sci, OMM, ONIX, Umachines, Agilent, Chorum, Corning, Kent, Optogone, OptXCon, Spectra Switch, Optun, ThreeFive Photonics, Civcom, Lynx, EOSPACE, Brimcom, LMG, LIGHTTech, Optical Switch, Trellis) or 3D (Active Optical, Agere, AIP, Blue Sky, C Speed, Corning, IMMI, Memlink, MEMScAP, MEMX, SiWave, TI, Transparent Networks, Continuum Photonics, Polatis, Creo)

(What a motley crew)
Page 1 / 3   >   >>
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE