OFC's Hot Products

Last year’s Optical Fiber Communications (OFC) conference went down in history as the place where Agilent Technologies Inc. (NYSE: A) announced its bubble-based optical switching fabric, sending its stock price through the roof (see Agilent Unveils Optical Switching Breakthrough).

It’s also remembered for what looked like the first monster MEMS switch from Xros (see Xros Launches First 1000-Port All Optical Cross Connect), the startup that subsequently got bought by Nortel Networks Corp. (NYSE/Toronto: NT) for stock then worth $3.25 billion (see Nortel Buys a Monster Crossconnect).

Both of those developments ended up not being quite what they seemed. Agilent has only named a single customer for its bubble technology, Alcatel SA (NYSE: ALA; Paris: CGEP:PA), and Alcatel appears to be hedging its bets, judging by its investment in OMM Inc., a manufacturer of alternative MEMS-based switching subsystems. And Xros’s monster switch turned out to be a midget in disguise (see Xros's OFC Splash Was All Wet).

So, what about this year? Which new products appeared to push forward the frontiers of optical networking -- and which of those are likely to live up to their promises?

Here are the ones that Light Reading editors got excited about, in random order:

Genoa Corp. unveiled its flagship product, which it calls a Linear Optical Amplifier. This widget is essentially a mutant semiconductor amplifier (SOA). It provides moderate gain (10 to 15 decibels) in a small, inexpensive package, so it could become popular in metro network applications, where cost and size are at a premium. Unlike other SOAs, Genoa's amplifier is suitable for multichannel transmission. The gain stays stable, even when channels are added, dropped, or switched -- that's another improvement on erbium-doped fiber amplifiers (EDFAs). To make this happen, Genoa has integrated a vertical cavity laser into the SOA to provide gain saturation. For more details see Genoa Amps Up.

Kamelian Ltd. demonstrated a reconfigurable optical add-drop multiplexer (OADM) with no moving parts. Inside the device, three SOAs control the path of the signal -- one for add, one for drop, and one on the pass-through. The SOAs act like switches. Their big advantage is that they have a very high isolation -- the difference between the on and off states is 50dB. As an added bonus, the SOAs also provide amplification.
Zenastra Photonics Inc. also deserves a mention for its reconfigurable OADM, which is based on thermo-optic switches (see Zenastra Launches Tier-2 Devices).

KVH Industries Inc. (Nasdaq: KVHI) announced ActiveFiber, a technology that the company says enables it to put components directly into optical fiber. The benefits of the technique, according to KVH, include improved signal quality and elimination of extra parts that would otherwise be needed to link fiber to components. KVH says it plans to release a 40-Gbit/s optical modulator based on ActiveFiber, but it hasn't said when the product will be available. See Optical Fibers Go Active for more details.
KVH isn't alone. Pacific Wave Industries Inc. announced a similar 40-Gbit/s fiber modulator at OFC (see Teledyne, Pacific Wave Partner Up). But Pacific Wave's design appears to be more complex than KVH's. The only true comparison of the two will come when products actually ship.

Luxcore Networks Inc. showed a working prototype optical switch that incorporated optical wavelength conversion -- thought to be a world first. This development promises to slash carrier costs considerably by eliminating expensive transponders. Optical wavelength conversion also means that Luxcore can make more efficient use of switching fabric without risking blocking problems. For details see Luxcore to Demo Optical Switch Advance. Downsides? This was a demo of a small 2x2 switch. Making a larger switch suitable for commercial use still presents an enormous challenge.

Lynx Photonic Networks demonstrated a working 4x4 switch made in lithium niobate. Insertion loss was around 6dB, remarkably low for this type of material. Its switching speed -- less than 5 nanoseconds -- means that the device could be used to link together existing high-speed switches and routers to make monster machines, capable of handling many terabits a second of data. See Lynx Launches NanoSecond Switch for details. The full product specification has been posted on http://www.lightreading.com/boards/message.asp?msg_id=4842

MetroPhotonics Inc. described its bidrectional mux/demux device. It's based on an echelle grating, which works in a similar way to arrayed waveguide gratings (AWGs). In an AWG, light is split into its constituent wavelengths by a fan of waveguides of incrementally varying lengths. In an echelle grating, a stepped mirror formed by a vertically-etched facet performs the same function. The echelle grating idea isn't new -- in fact, it was the precursor to the AWG. But when it first appeared, it was very difficult to etch the stepped mirror. Fabrication techniques have improved hugely, and MetroPhotonics claims that now it's possible to make echelle gratings that will outperform AWGs. Like AWGs, echelle gratings are integrated technology, but they are about a quarter the size of AWGs. MetroPhotonics has filed a patent on an innovation that gives its device some tolerance of wavelength variations in the pass band. As a result it can be used for both multiplexing and demultiplexing.

Novalux Inc. launched the first products in a family of high-power pump lasers. The lasers are based on vertical cavity surface emitting laser (VCSEL) technology, so they can be tested on-chip, which saves money in the manufacturing process. For a description of how the laser works, see Novalux Details Laser Advance.

OptiMight Communications Inc. launched an ultra-long-haul transmission system capable of supporting 1.6 Tbit/s over a single fiber without using Raman amplification. That may prove significant if Raman proves to create significant maintenance problems for carriers (see Raman Risks Emerge ). For details of Optimight's product announcement, see OptiMight Details Long-Haul Box.

Teem Photonics announced its erbium-doped waveguide amplifier (EDWA) gain block. It works in much the same way as an erbium-doped fiber amplifier (EDFA) gain block, but it is much smaller because the 20m spool of erbium-doped fiber is replaced by a waveguide just a few millimeters long. It's small, cheap, and easy to assemble, says Teem. For more details see Teem's Tiny Metro Amp Makes Waves.

Velio Communications Inc. unveiled a chip that promises to enable systems vendors to build much larger-scale optical switches capable of grooming thousands of STS1 (51.8 Mbit/s) channels. Right now, the biggest switch of this type boasts 512x512 ports, each handling 2.5 Gbit/s. Velio reckons it can more than triple this size. Its chips incorporate a 72x72 switch and 72 pairs of input/output devices on the same piece of silicon. One hundred and twenty of them are needed to make a 1,728x1,728 switch. For more details, see Velio Breaks Grooming Barrier.

-- Pauline Rigby, senior editor, and Peter Heywood, international editor, Light Reading http://www.lightreading.com
Page 1 / 2   >   >>
prolyxceo 12/4/2012 | 8:35:28 PM
re: OFC's Hot Products Hi DKP,

I agree with you about the difficulty in manufacturing DWDM components. However cost reduction will force outsourcing. We will see in the future what happens.

Thanks for the input.

farmboy 12/4/2012 | 8:36:26 PM
re: OFC's Hot Products sorry.
SectorInvestor 12/4/2012 | 8:36:27 PM
re: OFC's Hot Products <<at 'd'="" 28="" a="" alignment="" and="" as="" device="" every="" fiber="" from="" good="" have="" insertion="" issues="" loss="" manufacturter.="" ole="" optical="" or="" planar="" point="" polymer="" round="" same="" similar="" smf="" some="" still="" the="" their="" they="" to="" transition="" type="" waveguide="" will="" with="">>

I checked this out with the company. This is their response:

"The spec sheet indicates that the loss includes all connections in and out--including the coupling loss back into the standard telecom fiber."

Per the PDF document off their website at:

the TOTAL insertion loss is 4db on average and 5.5 at maximum. I think you will find those compare favorably, even to 10Gbps modulators, let alone 40Gbps where the losses are much higher.

So there is one advantage of the "D" fiber. Others are low drive voltage at 3V DC, a high extinction ratio of 23db, and a significant cost advantage. The combination of those advantages should be quite substantial for them.</at>
ComponentGuy 12/4/2012 | 8:38:56 PM
re: OFC's Hot Products Let me preface these comments that while I have no stock in KVH one can assume I would be percieved as a competitor to them, so take these comments with however many grains of salt you would like.
I would agree that polymer technology holds some potential cost savings as previously noted. What I'm not seeing the advantage of is the use of the 'D' fiber. I only see it as a replacement to the silicon substrate used in other designs which is not a significant cost. At some point they still have to transition from their 'D' fiber with a polymer waveguide to good ole round SMF 28 or similar type fiber and they will have the same optical alignment and insertion loss issues as every planar device manufacturter. Again I don't mean to discount the potential advantages of the use of the polymer waveguides, I just don't see the advantage of using the 'D' fiber, that appears to be a little marketing hype to me,imho.
tryit 12/4/2012 | 8:39:05 PM
re: OFC's Hot Products Can anyone comment on the total market size
of Velio's chip ? What will be the total
number of ports whole country will need.
pali 12/4/2012 | 8:39:20 PM
re: OFC's Hot Products anyone havning information abt key startups in packet over lambda, knowledge resources like whitepapers, univ work etc...
DKP 12/4/2012 | 8:39:21 PM
re: OFC's Hot Products

> We all know that there are too many
> players in the DWDM component market.

Yes. There were a gazillion of them at OFC.

> The proven method of reducing cost is
> outsourcing or contract manufacturing

It is hard to do with DWDM components:
(1) Fiber handling is still a skill (trained) labor task. Few contract manufacturers are trained to do this.
(2) If you outsource, you pay profit margin from the contract mfg, which eats into your profits.
(3) Off the shelf Manufacturing tools still do not exist for this process. Each DWDM design is slightly unique, and manufacturing tools are custom. If you look at Newport, etc tools, they all have to be customized to build your component, and it is a huge $$$$ investment for automated positioning and welding tools.

PS: One of the coolest products I saw at OFC was an Athermal AWG (no temperature control needed). That has to be the lowest cost component in the long run.


prolyxceo 12/4/2012 | 8:40:52 PM
re: OFC's Hot Products Hi All,

We all know that there are too many players in the DWDM component market. Cost will quickly become a factor and most of these companies will disappear shortly if they do not reduce the component cost.

The proven method of reducing cost is outsourcing or contract manufacturing, yet many of the start-up's have not realize that they must outsource there packaging.

Why do most of these companies continue with the strategy that manufacturing must be done internally?

Is it that there are few companies that provide contract manufacturing service to this emerging market?

Who are the suppliers custom packages to the fiber-optic market?

What are your thoughts?.

Indinvestor 12/4/2012 | 8:40:53 PM
re: OFC's Hot Products As an investor who is familiar with KVH Industries, I can assure you their product is not vapor ware or simply a presentation based on hype or empty promises.

KVH, relatively unknown to the commercial fiber optic sector, is emerging into the commercial communications arena after many years of product development and delivery for military and marine applications. They are no strangers to the precision required to produce durable products for dependable service. Their military and marine sales continue as divisions of the company today, but they are now expanding into the commercial applications of both new and established technologies.

I believe KVH will prove to be a reliable and worthy competitor in bringing the next generation of optical equipment into the mainstream. It could be a grave mistake to prematurely discount their potential to innovate new products simply because you may not have heard of them before now.

Build it. They will come.
SectorInvestor 12/4/2012 | 8:40:56 PM
re: OFC's Hot Products A lot of time and money is being spent to develop Polymer based electro-optic chips and to try to extend the reach of crystaline materials into 40Gbps.

The Polymer approach now appears to have major advantages in high speed, and low voltage.

Now KVH appears to be able to bypass the electro-optic chip entirely, using an all optical fiber approach. This will add significant advantages in insertion loss and manufacturing cost as well.

Their approach is protected because only they can make this special "D" fiber, which is heavily patent protected. They have other key patents and applications covering this fiber approach as well. They have over 20 years of fiber optic research and development to build on.

This combination of advantages, coupled with a definite push to get to 40 Gbps, will attract a lot of interest. KVHI was below the radar of the Optical Community for a long time, but not for much longer.

No wonder CheesyGuy appears worried.
Page 1 / 2   >   >>
Sign In