"We've been awarded a project, it's a long-haul application in a tier 1, and we do intend to announce it this year," CEO Tom Fallon said on Wednesday's earnings call with analysts.
And he does mean a U.S. tier 1. He'd mentioned a domestic win earlier, and that statement was a response to a question about it.
Infinera has hinted at this before, without disclosing the customer. Verizon has been the popular guess, mainly because the companies have hung out together so much. Stuart Elby, Verizon's vice president of network architecture and technology, was at last year's launch of the DTN-X, Infinera's next-generation optical transport system. (See Verizon Teases at Infinera Event and Is Verizon an Infinera Customer?; free registration required for that second link.)
Infinera has set itself a Dec. 31 deadline to tell us. The company also has a Wall Street analysts' day set for Dec. 6 in New York, as executives noted on Wednesday's call, so they'll even have a venue for anything big they want to talk about.
Heavy Reading believes that the optical industry will enter a new phase of photonic integration as it moves to 100G and beyond to the era of super chan-nels, as described earlier in this paper. Significantly, the large-scale PIC, by its design, fits nicely into a super channel model that requires multiple carriers/lasers.
Yes, I hear you. I shouldn't invest in tech, but every once in a while I buy on dips when I think things will tick back up. I call this my "gambling money..." and keep it to a VERY small % of my investments.
But, to your point, I would be wiser to not even invest those small amounts into tech. I would do better in other things as you point out.
> They have the 2 step problem.They have to get the PIC right AND they have to get the system right.
No question. My list is focused on high-speed only, and my last comment adds a third problem to your list: They need to make a profit, which has not exactly been their strong suit.
> Am I better off just putting more money in P&G.
The fact that potato chips have a better return than silicon chips is a bizarre fact of our era.
They have the 2 step problem. They have to get the PIC right AND they have to get the system right. That double investment will always plague them in today's first in winner game. Note: I am not doubting their technology or anything just their entire business model. I have been skeptical about that for years here. One misstep in any of those developments (which I have yet to see a team that NEVER makes a mistake) and poof they are in deep trouble.
seven
PS - Sailboat - one of my investing rules is NEVER invest in tech. Tech companies want huge P/E ratios and high PEGs for something that is either not disruptive or is not a winner on a large scale. I always ask this question...Am I better off just putting more money in P&G. :)
> Actually, the low latency trend has been a shift to wireless communication because it travels line of sight instead of contour of the Earth. You get about 800us improvement in latency in Chicago to New York by going wireless. But I'm sure you knew that.
And here I was thinking that the latency improvement had to do with fancy-dancy words like "index of refraction", but today I learned that it's because you can see NYC from Chicago on a clear day.
More generally: A repost, but worth noting the review of infn's high-speed record so far:
IFN gets only 100G per lambda using QPSK. No different than others in this regard. They don't have any majic modulation or detection scheme that gives them more bits than others per lambda. Their 500g PIC is using several wavelengths to achieve the 500G.
Either more lambdas to get to higher bit rates, or different modulation (like QAM) to get more bandwidth.
IFN does not have any advantage in the modulation area. Everyone is stuck by the same laws of physics.
Their advantage is tight integration of hardware, asics and software, and co-packaging several optical elements into a single hermtic package. Hence, in theory, lower costs and full regeneration at each node with PICs. (of course they also use EDFA's to extend their reach before regeneration).
Also note: there have been several papers published over the last few years showing QAM being transmitted long distances at high bit rates. So there are techniques to extend the reach of QAM.
QPSK only gave some of the gains in spectral efficiency by giving more bits per symbol.
Coherent Detection gave the rest. You can use coherent methods on QAM.
DNA, yes in a vacuum wireless has less latency. With line of sight through the atmosphere you only get the lower latency using wireless in fairly short hops as the atmosphere attenuates the signal. and wireless cannot scale to the bandwidths we need, so largly those long distance micro-wave based networks have been abandoned and upgrated to fiber links. Microwave of course is used heavily in wireless backhaul. For this application, with Low latency, sufficient bandwidth, and don't have to drag fiber to all the macro cell sites, wireless works well.
Infinera get "500G with QPSK" exactly the same way as everybody else -- by using multiple parallel 100G (or 50G?) coherent QPSK channels, each using 50GHz (or 25GHz?) bandwidth. The difference is that they use one PIC for 5 (or 10?) channels instead of a separate PIC (or discrete optics) for each channel.
But this single PIC then has to connect to multiple 100G ADC/DSPs to do the coherent reception, just the same as everyone else. Of course it's possible to integrate multiple channels of ADC/DSP into one device, but this then means more power to get out of one package which is not necessarily system-friendly.
For more than 100G per channel you can use 16QAM to get 200G into one 50GHz optical channel, then parallel up these to get 400G (or 1TB if you're feeling aggressive). Range is shorter (lower OSNR) but fiber capacity is doubled. But since lots of links (at least in Europe) are 500-1000km this is fine; for longer reach you still need 100G QPSK.
Infinera's PIC technology is undobtedly very elegant and will presumably give them a cost saving for multiple channels, at least compared to discrete solutions which are what is in the market now; whether it's really better (or cheaper) for building a system than multiple single-channel PICs remains to be seen.
The blogs and comments are the opinions only of the writers and do not reflect the views of Light Reading. They are no substitute for your own research and should not be relied upon for trading or any other purpose.
More Valley Wonk
Where SDN Is Going Next 6:00 AM Well, it seems to be going everywhere, but here are some guesses as to what might come of that
Juniper Cracking SDN Open 5:00 AM The $176M acquisition of startup Contrail looks like an attempt to create an open-source movement in software-defined networking
Photos: Broadcom's CES Preview 1:35 PM From an 18th-story San Francisco office, Broadcom shared its ideas for a wireless near-future
Buying Into the New Cisco 6:00 AM For all the talk of a software-and-services makeover, Cisco might not be ready to let the switch/router market fall prey to SDN
SDN & the Commodity Question 6:00 PM Software-defined networking won't eradicate Cisco or Juniper. At least not right away
To save this item to your list of favorite Light Reading content so you can find it later in your Profile page, click the "Save It" button next to the item.
