Optical components

ROADMs: Almost Famous

The metro networking market needs and wants reconfigurable add/drop multiplexers (ROADMs), though the products are likely to remain a niche play for now. But that niche is still an important part of U.S. RBOCs' metro growth plans, according to "ROADMs and the Future of Metro Optical Networks," a report issued this week by Heavy Reading, Light Reading's paid research arm.

ROADMs are the latest iteration of optical add/drop multiplexers. They are traffic cops of Wavelength Division Multiplexing (WDM) nodes, stopping some wavelengths for termination, allowing others to pass through, and adding new wavelengths to the stream. The result is an optical network that can more easily adjust to changes in traffic patterns.

The concept has been around for a few years now, and vendors at Light Reading's Future of Optical Networking conference in April noted that the technology is ripe for mass deployement (see Optical Networking: All Grown Up).

The problem is price. "The marketplace has long been rather wary of ROADMs" because of the price premium commanded and the fact that the products tend to incorporate new technologies, writes Heavy Reading analyst Scott Clavenna in the report. "The market seems to be in a 'chicken/egg' situation: The volumes necessary to drive down pricing will be difficult to achieve in a cautious spending environment."

Even so, ROADMs appear to be entrenched in carrier plans, with most service providers expecting to begin bulk deployments in 2006 or 2007, according to the report.

Recent triple-play initiatives provide a nice opportunity for ROADMs because they will lead to growth of metro networks in general. Along the same lines, cable network expansion is helping the ROADM cause.

Among the equipment vendors offering ROADMs, the report ranks Fujitsu Network Communications Inc. (FNC) and its Flashwave 7500 offerings at the top of the class (see Fujitsu Firms Its ROADM Resolve). Private vendors Movaz Networks Inc. and Tropic Networks Inc. scored among the leaders as well.

ROADMs are difficult to judge as a group, because so many approaches are offered, from older wavelength-blocker techniques to the more flexible wavelength-selective switch. Most ROADMs use an all-optical approach to adding and dropping wavelengths, but it's also possible to perform the function with a conversion to electrical traffic. Infinera Inc. and Meriton Networks Inc., both included in the report, exemplify this "OEO" method.

The report, available for $3,495, surveys 20 equipment vendors and includes feedback from 10 major carriers.

— Craig Matsumoto, Senior Editor, Light Reading

Page 1 / 3   >   >>
o-man 12/5/2012 | 3:14:15 AM
re: ROADMs: Almost Famous Does Lightconnect still make a blocker??? And did they qualify the 100 ghz and the 50 ghz???

Also Is the JDS 100 ghz blocker c-band, or just x-L band???

as far as I know the JDSU 100 ghz blocker is just for Lucent like stuff.
optigong 12/5/2012 | 3:14:15 AM
re: ROADMs: Almost Famous O-man -
I have to disagree. JDS, Lightconnect and Avanex have "qualified" Blockers. Xtellus qual was in progress a few months ago, it should close to complete by now. That's 3 1/2 suppliers to chose from.

On Capella, I agree, they still have to show a WSS that works for all the money they've spent. They might get there eventually if the VC's keep funding it.
o-man 12/5/2012 | 3:14:15 AM
re: ROADMs: Almost Famous How can WB's be old technology when systems houses are having trouble finding a "qualified" blocker?

Also has the pass band narrowing and insertion loss ripple problem been corrected for the "low cost" PLC based WSS to really be deployable?

Are we depending on a company like "capella", that is not PLC based, to be a success??? Please don't because from what I can see they have spent all their money 2 times and only have a hand full of "samples" to show for it!

Also I agree and think Fujitsu is in the lead but where is Marconi?
Balet 12/5/2012 | 3:14:13 AM
re: ROADMs: Almost Famous Lightconnect does not have a qualified blocker and people to finish the qual. All their sales is gone.
Xtellus can not finish a qual, may 50GHz L-band...
JDSU has not qualified their LC blocker yet. They sell it because nobody else can make it and survive.
Avanex is looking for the blocker to make Tropic/Alcatel happy. They will never qualify their crapy product.
Coadna is in deep...

It means, that there is no qualified blockers exist. Everyone is moving to WSS, but can't make them well yet.
o-man 12/5/2012 | 3:14:12 AM
re: ROADMs: Almost Famous Balet Wrote: "It means, that there is no qualified blockers exist. Everyone is moving to WSS, but can't make them well yet."

I respectfully disagree. Fujitsu's ROADM is in the lead...right? They are using the switches and VOA's...I think... The blockers when will in the short term because they are available. WSS's are really not ready yet otherwise they would be shipping in large volumes...

I heard a rumor that the giant JDSU can only make 1 WSS per week!! haha
Fyio 12/5/2012 | 3:14:10 AM
re: ROADMs: Almost Famous CoAdna has qualified blockers and deliverred triple-digit volumes:


They also introduced WSS at OFC/NFOEC'05:


dc_optics 12/5/2012 | 3:14:10 AM
re: ROADMs: Almost Famous Cisco is using the WSS in there 32 channel 15454 DWDM boxes.

I heard that they have been shipping these cards out for the past nine months to several large SP's.
Mako 12/5/2012 | 3:14:09 AM
re: ROADMs: Almost Famous Given the questionability over WB qualification I wonder how the industry ever expects to see qualified WSS given greater technical complexity, and smaller market space (mainly degree 4+ nodes for ring interconnection).
Balet 12/5/2012 | 3:14:09 AM
re: ROADMs: Almost Famous It is a BS.
Qualifying a WDM module against 1221/1209 which are designed for passive components is a BS.
I was also told by their potential customers that they are half-existing.
I was also told by their internal source that they have a great liquid crystal team, but poor or no engineering and S&M.
I wonder if Jim can speak up if it is true or not.
flying4s 12/5/2012 | 3:14:06 AM
re: ROADMs: Almost Famous Does this conversation make sense?

At best there is a claim that there are 500wb's (more likely 300) required per year. Assume that someone actually got smart enough to make a $1 off it, heck let them make a $1000/unit against a what $10,000 - $20,000 selling price, and holy smokes you just made $0.5M. Now here's a viable business! Get real, the only one that can play in this type of market is someone that uses wb as an element in an established product line. Ugh, JDSU.

So lets split the $0.5M bonanza over the folks playing the game and see what happens:
Lightconnect - bagged out a couple of months ago, couldn't get their act together, how much did they lose in the process?
Xtellus - chuckle, sure a "player". I bet through struggling mightly they now have a dominant 3% of the market (that's $15,000! time to start making ice sculptures!)
Avanex - have you seen one of these? Yeah they're selling but it's based on who they know.
JDSU - The usual first supplier, supposedly qualified, failing in the field, customer's scrambling for a second source
<add here="" more="" names="">

Now look at the customers.
Lucent - hah, who's going to be the poor fool that will try to sell here? I think most of the vendors have learned their lesson by now.
Siemens - Why? Are they shipping something?
Marconi - Listen closely, that's taps you hear in the background
Huaweii - Sooner or later they'll find someone to supply below cost product until the vendor goes broke
Alcatel - Who are those guys? Do they have an office?

So now with such a great market and a healthy supply of competitors let's upgrade! Add the WSS, and since most of the vendors died on the wb then let's start using new players.

Capella - ROTFL. Come on, at best they are still a year away from knowing what's wrong with the "samples" they are selling.
Metconex - If it works, then wait for it to destroy the company and then sell it to...
JDSU - Ah, yes here we go again.

I repeat

Does this conversation make sense?</add>
Page 1 / 3   >   >>
Sign In