Movaz on the Move

Optical switch startup Movaz Networks broke cover this week. But the company raised many questions along with its emergence.

Movaz surfaced by signing a ten-year agreement with The BF Goodrich Company (NYSE: GR), which calls for Goodrich to supply MEMS (micro-electro-mechanical systems) exclusively to Movaz. The startup's product is expected to start beta tests later this year.

The announcement gives a general idea of what Movaz is up to. "We're building a platform capable of delivering wavelengths at the same price and in the same way STS1s are delivered today," says Movaz CEO Bijan Khosravi.

To make this happen, he says the product will incorporate its own management and signaling for setup and tear-down of wavelengths using DWDM (dense wavelength-division multiplexing).

Movaz's interest in this type of optical signaling is evidenced by its participation in the Generalized MPLS (multiprotocol label switching) working group at The Internet Engineering Task Force (IETF).

Indeed, Movaz places sufficient priority on standards work to have named one of its executives, Lou Berger, to the post of VP of protocol development. Berger, a former consultant and IP expert, helped coauthor the G-MPLS signaling draft issued September 2001. According to that document, G-MPLS "describes extensions to MPLS signaling required to... [extend] MPLS control... to encompass time-division (e.g. SONET ADMs), wavelength (optical lambdas) and spatial switching (e.g. incoming port or fiber to outgoing port or fiber)."

Movaz's participation in the IETF signals not only that it aims to conform to IP standards, but that its product will operate in peer-to-peer fashion across optical networks. (See Top 10 Trends for a clarification of how this differs from other proposed approaches to "intelligent" optical networking.)

Khosravi confirms this. "Our product will operate in distributed fashion in an optical network," he says. "It won't rely on a central switch."

Given all these factoids about Movaz, it's likely its switch will fit into the optical market alongside other MEMS-based optical-electrical-optical (OEO) crossconnect switches from the likes of Brightlink Networks Inc., Ciena Corp. (Nasdaq: CIEN), Cisco Systems Inc. (Nasdaq: CSCO), Nortel Networks Corp. (NYSE/Toronto: NT), Sycamore Networks Inc. (Nasdaq: SCMR), Tellabs Inc. (Nasdaq: TLAB; Frankfurt: BTLA), and Tellium Inc..

Marketing Position But questions remain about whether Movaz can execute on its plan. For one thing, what about Goodrich? On the plus side, Movaz has cleverly aligned itself with a larger manufacturing partner that not only can provide it with capacity and parts but also needs its help. Goodrich, having divested itself of various businesses such as the making of tires, is eager to emphasize its identity as an engineering company with cutting-edge expertise. Getting involved with an optical startup showcases this transition.

Goodrich does have years of MEMS experience in aerospace and automotive applications and a 15,000-square-foot MEMS R&D facility with 85 engineers in Cincinnati. Still, it remains to be seen how well Goodrich can make the shift to telecom applications.

Movaz has already scored $30 million in funding from Menlo Ventures and WorldView Technology Partners. It has over 180 employees. Headquarters are located in Norcross, Ga., with offices in Tyson's Corner, Va. Movaz also will be sharing Goodrich's facilities in Cleveland on a permanent basis. And plans are underway to open offices in Ottawa.

The executive team includes Khosravi, who headed up marketing at Siara Systems before and shortly after its merger with Redback Networks Inc. (Nasdaq: RBAK) (see Redback's Khosravi Question Answered). Daniel Awduche, most recently of WorldCom Inc. (Nasdaq: WCOM), is VP of network architecture. David Smith, a veteran of Bellcore and a former professor at Ohio's Case Western Reserve University, is CTO. Zouhier Mansourati, formerly of Nortel, is VP of product line management.

-- Mary Jander, senior editor, Light Reading http://www.lightreading.com

skeptic 12/4/2012 | 8:43:10 PM
re: Movaz on the Move
Its important to understand that there is not just one set of standards (or one standards body) that is relivant to the sort of product they are talking about. And the industry overall is far from any sort of consensus on generalized MPLS as the right solution to this problem.

movaz will be hurt as much as helped by its affiliation to certain standards bodies. And its ability to respond to change will be limited by its close alignment at the beginning of the product cycle to a standards direction that may or may not work out.

Dr. Dre 12/4/2012 | 8:43:09 PM
re: Movaz on the Move Good point. I am also curious to know how many vendors intend on being fully GMPLS compliant.

Although there seems to be a concerted effort amongst vendors to agree on a common framework, I wonder whether some will abandon ship and go there own way (a la Sycamore, Corvis)?

Or maybe, in the end everybody will just do whatever Cisco does?

alloptical2000 12/4/2012 | 8:43:08 PM
re: Movaz on the Move
Your concerns are valid.

Butone of the biggest challenges is to be able to take that risk in associating to one standard, as they can not afford to all the standards bodies, due to scarcity of resources. This, as far as i know happens with each and every start-up.

I am not if this is true, but some times jumping on to a specific standard helps. I think this worked for Juniper on MPLS standards, though it did not work for sycamore on the ODSI for control plane signaling in the optical domain.

Do any one of you ahev any feedback on to the which standards body is most promising to get approved or standardized. The ones i know competing head-to-head are GMPLS(IETF), ASON, ODSI.

any insight or comments are appreciated.

xoip 12/4/2012 | 8:43:08 PM
re: Movaz on the Move Can some explain if this layer 2 device can impact viability/ need for high end routers like the junipers and GSRs.

Thanks in advance
Dr. Dre 12/4/2012 | 8:43:03 PM
re: Movaz on the Move The success of a standard depends on how many vendors implement it. That doen't bode well for ODSI since, only a few of the big guys were involved. In fact, one such "big guy" (Ciena) has recently become quite involved in producing IETF documents.

The ODSI framework isn't nearly as sexy as GMPLS, but somehow, it is the one that seemed to make the most practical sense to me (at least in the near future) because it mandates that the client domain be seperate and distinct from the optical backbone.

As for ANSI's AON, I don't know much about it. I did take a look at some works posted in the T1 website (www.T1.org) a year ago, and all they seemed to be doing was defining some goals and objectives. Frankly, I think ANSI works at such a slow pace, its almost laughable.

The IETF group seems to be getting bigger by the day. This isn't necessarily a good thing. While several brains are indeed better than one, too many cooks can spoil the broth. The IETF meetings are notorious for infighting between various vendors.
ng_mui 12/4/2012 | 8:42:47 PM
re: Movaz on the Move You can forget ODSI, the main proponent was Sycamore, with some other vendors (6 or so)
involved in interoperablility tests, but
basically the work is finished.

The most promising activity is happening in the
IETF. There is similar work going on in the OIF,
but mainly this is an implementation body trying
to drive the adoption of the IETF drafts.

As far as the ASON work is concerned, it's mainly
architectural work from the standards bodies -
in other words, forget it until 2005!

As other posters have mentioned, the GMPLS work
in IETF is immature with interoperability under
question (which is why it's being addressed in
OIF) - for example there being two completely
different sets of signalling protocols!

Basically the idea is to interconnect the optical
transport network (e.g. an edge optical NE)
with a core IP router. The cynical view is that
the optical switch manufacturers will implement
whatever Juniper and Cisco say!

It's still an open point how the new signalling
will work together with existing centralised
network management.

wOOp 12/4/2012 | 8:42:46 PM
re: Movaz on the Move Anyone who thinks standards influence or affiliation is critical to their success are incredibly naive.

Just look at MPLS and now G-MPLS. It has become a bloated bunch of junk trying to do everything to everybody. Can we say ATM all over again.

By looks of the numbers attending standards meetings, I'm amazed how many engineers obsess over standards as if participation is the path to engineering achievement.

Gag!!! Most standards attendees and authors couldn't generate a working C++ or Verilog module if their life depended on it. And those that think they can are obsessed with creating the most complicated thing known to man (probably because they don't have to implement it).

If Movaz thinks standards are key, they are going no where.

Standards are only a necessary evil, not the central core of a business strategy.
netskeptic 12/4/2012 | 8:42:45 PM
re: Movaz on the Move > Just look at MPLS and now G-MPLS. It has become
> a bloated bunch of junk trying to do everything
> to everybody. Can we say ATM all over again.

There is a lot of junk in ATM too, however, it seems like say born-again-ATM or ATM-with-human-face would have a very good chance to be the next generation data-link protocol of choice, however, it is not going to happen as long money are being wasted on various vaporware.


Sign In