Infinera introduces Infinite Capacity Engine

Investment Analysis Clubs Saul’s Investing Discussions
1

And here it is:

Sunnyvale, Calif., – March 21, 2016 – Infinera, provider of Intelligent Transport Networks, announced the introduction of the Infinite Capacity Engine, a multi-terabit optical subsystem enabling a step function in optical wave division multiplexing (WDM) performance. The Infinite Capacity Engine is the first optical subsystem in the industry to offer network operators the combined benefits of delivering optical super-channel capacity up to 2.4 terabits per second (Tb/s) and reach up to 12,000 kilometers in a single small package.

The Infinite Capacity Engine is powered by the advanced electronics of Infinera’s next generation FlexCoherent® Processor and the cutting-edge photonics of Infinera’s fourth generation photonic integrated circuit (PIC). The Infinite Capacity Engine will be integrated into Infinera Intelligent Transport Network platforms customized for long-haul terrestrial, subsea, metro and data center interconnect networks to provide a comprehensive set of solutions to customers.

Transport network bandwidth requirements are growing exponentially, driven by the rapid growth of cloud-based services, increasing broadband access speeds, 5G mobile data, the Internet of Things and on-line video. To support the growing demand for on-demand network capacity, groundbreaking innovations are required.

A Step Function in WDM Performance

Enabling a step function in optical networking performance, Infinera’s new Infinite Capacity Engine is designed from the ground up to deliver an unmatched 2.4 Tb/s of optical bandwidth and long reach in the world’s smallest subsystem. In addition, the Infinite Capacity Engine consumes relatively little power – 82 percent less power per gigabits per second (Gb/s) than the nearest competitor’s solution.

The Infinite Capacity Engine delivers multiple industry breakthroughs:

Massive scale: Infinera’s Infinite Capacity Engine enables massive network scale supporting super-channels up to 2.4 Tb/s from a single subsystem with up to 12 times more bandwidth than other implementations.

Advanced Coherent Toolkit (ACT): The Infinite Capacity Engine delivers up to 60 percent more capacity-reach performance for submarine and terrestrial networks than the current generation of optical technologies by allowing extended reach and/or utilization of higher order modulation formats on a per channel or super-channel basis in challenging fiber environments. The Infinera ACT features Nyquist subcarriers, a new technology that increases tolerance for non-linear effects and SD-FEC gain sharing. In addition, multiple modulation formats are supported including matrix enhanced – phase shift keying (ME-PSK), BPSK, 3QAM, QPSK, 8QAM and 16QAM.

Sliceable super-channels: The Infinite Capacity Engine provides the unique ability to deliver 2.4 Tb/s of capacity in a single module with a single fiber pair. In addition, it provides fine-grained and flexible control of optical wavelengths at 100 Gb/s. The Infinite Capacity Engine can tune and route a 100 Gb/s wavelength or N x 100 Gb/s super-channels in multiple separate directions with any specified flexible grid coherent modulation across the extended C-Band. As a result, network line-card module requirements can be reduced up to 77 percent while delivering tremendous flexibility. Moreover, total cost of ownership can be reduced by up to 53 percent as compared to non-sliceable implementations.

In-flight wire-speed encryption: The Infinite Capacity Engine encrypts traffic transparently at Layer 1 on a per service basis or in the bulk mode per super-channel, using the strong 256-bit Advanced Encryption Standard (AES) key. The Infinite Capacity Engine is integrated with easy-to-use key exchange and cryptography mechanisms, providing wire-rate multi-terabit processing with more capacity than current industry solutions.

Enables New Flexible Network Architectures

With the Infinite Capacity Engine, Infinera has defined a new approach to deliver multi-terabit capacity in optical networks. The Infinite Capacity Engine enables the pre-deployment of bandwidth that is service-ready and can be provisioned on-demand in 100 Gb/s increments with simple software activation that leverages Infinera Instant Bandwidth. Instant Bandwidth has proven to be a powerful approach to rapid deployment of services, enabling new capacity to be deployed in less than one hour. Infinera’s standard license offerings have been expanded to include Time-based Instant Bandwidth providing 100G temporary licenses enabling capacity to be deployed for a specific duration of time. This capability, combined with up to 2.4 Tb/s of pre-deployed capacity, can be SDN-controlled in 100G increments, with each slice configurable for color, modulation and direction, which provides network operators with a way to dramatically reduce operational rigidity and forecasting complexity. The Infinite Capacity Engine allows network operators to instantly deploy massive increments of line-side bandwidth when and where they need it; whereas, it could take weeks or months to deploy bandwidth with a competing solution.

Press release:

https://www.infinera.com/next-step-function-in-optical-netwo…

Product home page:

https://www.infinera.com/technology/engine/

Video explanation (very worthwhile):

https://youtu.be/NaSfRJSekGs

Best,
–Kevin

24 Likes
2

Appreciate the information and links Kevin!
Thanks!
Fool on,
~Chuck

3

Hi Kevin,

Good detective work last week on figuring out when the announcement for this Infinite Capacity Engine would be! This morning, right on cue.

I posted this on the INFN board also, but thought some here might also benefit from the answer to the same dumb question I have.

I watched the video, and yes, it explains what this new engine is and does well. My question is (and frankly, one that I’ve had for a while), if Infinera is so far ahead of the competition as it sounds like they are (in capacity, low power usage, and low cost of ownership), why don’t they have a larger market share? It seems to me they ought to be the only brand worth buying!

Mike

2 Likes
4

if Infinera is so far ahead of the competition as it sounds like they are (in capacity, low power usage, and low cost of ownership), why don’t they have a larger market share?

A quick explanation would be because of existing relationships. They are competing against well-established players in the industry. Both Cisco and Ciena have been around since the 80’s and 90’s. Infinera developed their technology and approach during the 2000s and even operated as a stealth company for a bit. Their earlier incarnation was called Zepton Networks.

What will change that dynamic is 1) evidence of the market share they are gaining, 2) education and 3) the factors affecting competition among the buyers. Certainly with this news more buyers are getting an education and to be competitive themselves, they had better be considering the Infinera solution.

There’s something else, too.

The ability to break up light into smaller channels - those Nyquist sub-carriers - is really huge.

Normally (through optics) the data sent back and forth over long pipes needs be aggregated first so it can be sent across those long stretches. Once aggregated, the data then goes across the fibre optic cable along a given single carrier wavelength. Then, once it hits the other side the data must be broken down (disaggregated) so it may be distributed.

Infinera has found a way to essentially keep those wavelengths “disaggregated” all the way along in the pipe. They do this through the use of subchannels - Nyquest sub-carriers. This means less expensive equipment and less complexity on the upstream side for aggregating and the same on the downstream side. It also allows them to route and re-route information to various places before going through another aggregation/disaggregation process. It is hugely efficient and tremendously flexible.

The discovery on breaking down light into various subchannels isn’t new. It was studied before and there were difficulties - largely due to signal regeneration requirements along the way that increased the equipment cost at various “hops”. The result was the benefits attained from the subcarriers became a wash with the added equipment needed.

Infinera has solved the conundrum problem. What they have achieved this time isn’t just due to their PIC advantage anymore (although the capabilities of the PIC do lend to the solution) it is a legitimate IP advantage.

This new development gives them a very real capacity and management advantage to add to their existing advantage in power, performance and distance. With developments like this, Infinera should continue to chip away at Cisco and Ciena’s business.

Best,
–Kevin

11 Likes
5

Hi Kevin,

First I’d like to add my thanks to the many others regarding your excellent coverage and sleuthing regarding Infinera.

Secondly, do you mean by IP that they actually have a protected patent regarding their solution to the light subchannels problem? If that so that seems fantastic…rather than just having a trade secret with their PIC, which is under threat at being found out by their competitors at any time, with no warning, do they now have another game changer that is a little more secure and robust?

6

Infinera has found a way to essentially keep those wavelengths “disaggregated” all the way along in the pipe. They do this through the use of subchannels - Nyquest sub-carriers. This means less expensive equipment and less complexity on the upstream side for aggregating and the same on the downstream side. It also allows them to route and re-route information to various places before going through another aggregation/disaggregation process. It is hugely efficient and tremendously flexible.

I wonder if there are advantages (of slicing data packets into smaller chunks) in addition cost and flexibility. Could there be advantages of better security? Data is encrypted and then sent. If the data packets are larger then the customer A’s data needs to be aggregated, encrypted, and sent in a packet with other customers’ data. This co-packaging/encrypting of data would be like flying in a commercial airliner with potentially many other passengers. If the transport vehicle is smaller then it can make sense for customer A to have his own wavelength, sort of like a private jet. I’m no expert on data security so perhaps an expert can weigh in on whether smaller data packets can enhance security.

Chris

7

Data is encrypted and then sent. If the data packets are larger then the customer A’s data needs to be aggregated, encrypted, and sent in a packet with other customers’ data.

Hmmm. I would have expected the encryption to occur before the data is put into packets, aggregated or not.

8

Hi Billy,

First, and I think I mentioned this, the Nyquist subcarrier approach is not a new thing in and of itself, and thus it isn’t protected by a patent. To work, however, the signal needs be re-amplified along the way at multiple shorter intervals causing the need for additional equipment. The flexibility advantage gained through Nyquist was a wash because the economics weren’t worth it.

But Infinera found a way to keep the signal in tact without the degradation. I honestly don’t know how, but apparently solving this problem has been a topic of major study over the past several years. Their ability to overcome the bottleneck is more a testament to their DSP expertise. Will others figure it out? Probably. Will it matter? Probably not, or rather probably too late. I’ll try to explain.

The biggest beneficiaries to solving this problem would be for those who have technologies that can transmit speeds in larger bandwidth blocks. Currently, Infinera has a line speed of 500GB/s. The next competitor just got to 400GB/s. But, as Andy (buynholdisdead) eloquently pointed out to us, competitor solutions that desire higher line speeds need to “fill” with more line cards to fully take advantage of those speeds. I am not an expert here, but it is something you need to do to support and disaggregate to the appropriate number of client-side cards. It is a limitation of the competitor technology solution - all of them - by not having your components integrated into one monolithic design and burned onto a PIC. At least that is how I understand it.

It is really the pairing of this technology with the PIC that makes it a lot more interesting.

With this technology, you don’t need more fiber lines coming in and you don’t need to fill more lines cards in the chassis just to get what you need on the other side. You need just one line coming in for up to 2.4Tb of bandwidth - because - that bandwidth is already coming in sliced. It is the equivalent of 24 100G line cards coming in (or 6 400G cards if you’re the next competitor). And there is no fuss to fill in on the line side just to get what you need on the client side either, at least not for Infinera. If I were in Infinera’s shoes I’d say this is pretty awesome sauce.

I hope that explanation helps. And this quote, too, if you are worried about a competitor on the heels of catching up:

Analysts DCD spoke to confirmed that Infinera has a two-to three year lead on the rest of the optical industry with this development. Those competitors include Ciena, Alcatel-Lucent (now Nokia) and Huawei which allegedly is the closest one behind Infinera.

Best,
–Kevin

9 Likes