Why is this important? Because telecoms operators’ networks are always facing a capacity crunch, mainly thanks to the explosion in cloud and online video, and there’s always a need to pump more bits through the system. We’re now looking at the next generation of such connectivity, alive and kicking.

The link, which is based on Alcatel-Lucent’s 400 Gbps Photonic Service Engine, runs between Paris and Lyon, and the first tester is the French educational and research network Renater. The early use cases for this bump up from now-standard 100 Gbps wavelength technology will most likely be found in business and research, for services such as telepresence that will make good use of the boosted bandwidth.

“This link transports the bulk of France’s scientific data that passes through our network,” Renater MD Patrick Donath said in a statement. “This pilot phase also aims to test the latest switching equipment supplied by major OEMs on a network running at this capacity and will enable us the anticipate the architecture of Renater’s network in the coming years.

“A 400 Gbps network is an important step forward for the networks and research projects of tomorrow.”

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Cloud computing infrastructure: 2012 and beyond
• Mobile 2012 and beyond
• Facebook’s IPO filing: ideas and implications

What IBM has done is place a channel capable of guiding light on an everyday chip with the addition of only a few steps and equipment in the manufacturing process — speeding up the potential communications on the chips and even between chips. It’s also chosen to build the chips at 90 nanometers, an older technology that means fabrication plants will have the capacity to make these new semiconductors. Making a semiconductor involves several steps where materials are deposited on a wafer and then etched away according to a set pattern. It’s like building a layer cake with equipment that can costs hundred of millions of dollars.

But this equipment is already there and even some of the most expensive elements of it — the masks that set the patterns of the chips — are set to stay the same under IBM’s breakthrough. Solomon Assefa, a researcher at IBM, says that putting the nanophotonics on the chips will require the additional of germanium as well as deeper box on the chip where the optical components will reside.

However it still uses the conventional CMOS manufacturing which lowers the cost of manufacturing on-chip optics and helps make the chip suitable for applications such as the data center. In 2010 IBM piloted this technology and now it is ready to actually start building the chips. From the release:

“IBM’s CMOS nanophotonics technology demonstrates transceivers to exceed the 25Gbps data rate. In addition, the technology is capable of feeding a number of parallel optical data streams into a single fiber by utilizing compact on-chip wavelength-division multiplexing devices. The ability to multiplex large data streams at high data rates will allow future scaling of optical communications capable of delivering terabytes of data between distant parts of computer systems.”

Assefa couldn’t share IBM’s go-to-market strategy around the chips but expects the technology to handle everything from on-chip networking in high-performance computing and in big data processing workloads as well as speeding communications between data centers. That’s a pretty wide range of partners that IBM would work with to commercialize this technology.

Others are also trying to develop technologies here from Kotura and Luxtera to Lightwire, which was purchased by Cisco. Other companies, aren’t waiting for on-chip optics and are embedding fiber into top-of-rack switches. Using pulses of light to send information means more energy-efficient networking gear, no need for interconnects and much faster speeds that eliminate current bottleneck in higher performance computing and data analytics applications.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The importance of putting the U and I in visualization
• A near-term outlook for big data
• Dissecting the data: 5 issues for our digital future

These giants are deconstructing the server, rethinking the data center and building new databases. And they are also rethinking how they build out networks. To that end several companies have latched onto the concept of software defined networks to help build an abstracted and programmable layer on top of the physical networking gear. This gives systems administrators more flexibility, but it doesn’t help with another problem that many data center providers face, creating faster equipment that can carry their escalating traffic.

This is where Plexxi, a Cambridge, Mass.-based startup sees its opportunity. It has built what it calls a switch, but is actually an entirely new type of networking gear aimed at delivering a lot of traffic between racks of servers as fast as possible. Plexxi also offers Plexxi Control, management software that runs on commodity hardware and routes traffic around the network. Inside the Plexxi “switch” are optical components normally found inside telecom gear. Plexxi has designed the first networking gear for the data center that uses optics instead of electronics to solve the networking bottleneck.

Each switch has 32 10-gigabit-Ethernet capable ports and can be connected to other Plexxi switches. For now the company has tested deployments of its gear supporting 10,000 servers, but Plexxi CEO and CTO Dave Husak assures me it can scale bigger. It will have to. For example, Facebook, a potential customer, is estimated to have more than 150,000 servers. The Plexxi gear allows for the network to become essentially flat — a rack of servers talk to the box and the boxes have the speed and capability to talk directly to each other. Instead of hierarchies or trees — software running on the switch tracks where virtual instances have moved and routes traffic to them over this single layer. This speeds up the networking and simplifies the gear supporting the network.

Plexxi has done something unique in its box worth looking at. It has replaced the traditional electronics networking with optics, which boosts speed but also reduces costs and power draw. To keep costs in line it buys off-the-shelf components and strips them down to the essentials. While the transceivers and equipment inside telecom networks cost thousands of dollars, Plexxi’s entire box sells for $64,000 in part because it did away with things like line cards and amplifiers that telcos use to boost the distance an optical signal can travel.

Using optics inside the boxes and connecting the boxes directly together in a flat network also eliminates a lot of cabling and additional electronics that both add costs and increase the power consumption. Husak told me the Plexxi switch draws milliwatts per port. The latest switch from Arista draws 5 watts per port.

Husak is a veteran on the networking world, remembering how the switch replaced token ring and other networking architectures in the 90s. Now, twenty-some odd years after the first switch was unveiled, Plexxi is poised to remake networks again with an entirely new systems architecture and design. It has almost $50 million in funding and has been building this switch out since 2010. This is a company unafraid of taking on Arista, Cisco and other giants in networking in a quest to re-imagine the data center of the future.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Infrastructure Q3: OpenStack and flash step into the spotlight
• The promise of SDNs in the enterprise
• Cloud computing infrastructure: 2012 and beyond

The test used Infinera’s technology across 800 kilometers of a special type of fiber called Dispersion Shifted Fiber. This DSF Fiber is used in Japan, but currently can’t handle some of the advanced techniques companies are using to add more capacity to existing fiber. From the release for those who like the technical bits:

DSF fiber, deployed throughout Japan, has proven to be a challenge in deploying high-capacity dense wavelength division multiplexing (WDM) transport technology due to low chromatic dispersion in the C-Band, making it difficult to achieve high capacity optical transmission over long distance economically. As a result, capacities have been severely limited and have had to utilize expensive L-band technology to achieve long haul quality performance from the DSF fiber.

When using the same Infinera gear over modern fiber strands, the equipment can deliver 8 Tbps over distances of 2,500 kilometers, showing that even certain types of fiber aren’t as future proof as we might think, given our insatiable demand for bandwidth.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• GigaOM Research highs and lows from CES 2013
• How HR can make the case for workforce analytics
• The 2013 task management tools market

The £530m ($843m) rollout will cover what is known as the ‘final third’. Basically, telcos such as BT are perfectly happy to bring superfast broadband to two-thirds of the country on the basis that they’ll get a good return for it. For the remainder, they need a bit of a push, and that push is being carried out by a government department called Broadband Delivery UK (BDUK).

“BDUK… will assist local granting authorities in designing and implementing successful broadband support measures in line with EU competition rules,” competition commissioner Joaquín Almunia said in a statement. “The umbrella scheme will be a big step towards the achievement of the EU Digital Agenda targets and a strong impetus for growth in the UK.”

The biggest target this will help the UK hit is that of everyone in the EU having access to 30Mbps connectivity by 2020.

“Finally getting the green light from Brussels will mean a huge boost for the British economy,” UK culture secretary Maria Miller said in response. “Superfast broadband is essential to creating growth, jobs and prosperity and the delay has caused frustration within government. Today’s announcement means that we can crack on with delivering broadband plans, boosting growth and jobs around the country.”

Now, about that delay. As Miller points out, the European Commission took a few months longer than expected in giving its approval. The reason for this, most observers surmised, was the slight issue that almost all the government cash is likely to go to one company.

Indeed, in most of the UK, only two companies were left standing for the multitude of local tenders. BT was one and the other was Fujitsu – but Fujitsu messed up so spectacularly in previous public sector projects (particularly in the big National Health Service IT overhaul) that it’s now effectively blacklisted.

That leaves BT. It’s not as though BT will simply be sucking up money here – it is going to put its own cash into the final third deployments too – but it’s still a case of state aid benefiting one firm pretty much exclusively.

Almunia seems to think transparency and regulation will keep things on the level. The Commission statement reads:

“The UK telecommunications regulator [Ofcom] will have a crucial role in designing wholesale access prices and conditions. All information related to projects under the scheme (including mapping, public consultation, tenders, aid beneficiaries) will be published on a central website.”

For more detail, we’ll have to wait until the Commission publishes its reasoning in full. Either way, it’s full steam ahead for proper rural broadband in the UK – and, given that the first generation of copper-based broadband performed pretty awfully in those areas, the shift can’t come too soon.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Will Standardizing the Cloud Cause Clarity or Confusion?
• Google and the Ghost of Silicon Valley Past
• Netflix may suffer from limited mobility

Yes, both companies are deploying fiber and both will offer gigabit speeds, but that’s about the end of the similarity. Let’s start with the scope of the projects. Time Warner Cable is spending $25 million to connect “hundreds of buildings” in NYC, which means the cable company will extend its existing fiber to the building. At that point those tenants in the building will have to connect to the fiber in the building and bring it to their floor/offices. Analysts estimate Google is spending between $500 million and $800 million to connect parts of Kansas City. It’s not just the spending that’s different, and understanding what else is can help explain why the U.S. broadband infrastructure is not keeping up.

This is an evolution of TWC’s network, not a revolution.

Time Warner Cable isn’t providing a map covering the scope of the deployment, but it’s smaller in actual distance and in ambition than what Google is doing. When I asked for details, a Time Warner spokesman emailed me to say:

This will cover core fiber backbone build out and strengthening, as well as build out to individual buildings. With this investment we will be able to reach multiple hundreds of buildings throughout NYC. (Brooklyn, Long Island City (LIC) and other communities in Queens, major office areas in Manhattan such as Flatiron district, the Empire State Building, the World Trade Center, etc, and in select areas of Staten Island)…

However, even in its press release announcing the investment, Time Warner Cable noted that the company has just finished deploying fiber to The Empire State Building. So there’s a lot of spin in this news about the investment. Clearly some of these investments have been planned and are even already implemented, making this $25 million investment look more like an evolution of the TWC network rather than some gigabit fiber revolution.

The spokesman also specified that the Brooklyn Navy Yard and Greenpoint will get fiber to the building as well, and there is clearly an expansion of TWC’s service to those communities. However, Google’s investment is entirely new, something that happens only rarely. Time Warner Cable, like other networks, is always gradually expanding its fiber closer to the end user. That’s great, but it’s incremental and aimed in an area where it competes heavily with Verizon and Cablevision, so to avoid these investments would mean lost business.

Density and customers are key.

Another point of difference is that fiber to the building in NYC is different from fiber to the home in middle America. Google’s fiber to the home will give a household (the American average is 2.59 people) a gigabit to share. The TWC investment will deliver a gigabit to a building, where thousands may work and hundreds of customers might tap into the network. When it comes to deploying fiber, density lowers cost, and there are few places in the U.S. that are denser than New York City. That lowers the investment required to deploy the network, especially when one considers that most of the network has already been built out.

The final point of differentiation is on price and the actual market. Google is charging consumers $70 a month for access to its gigabit network (and $120 if customers want TV with that) while TWC is going after business customers. Those customers generally pay at least five times the price of consumers or more depending on the services they want. That makes the economics of offering business-class service very different — and generally makes it easier to justify investment.

So, higher prices (although also higher service levels), greater density and the evolutionary as opposed to revolutionary nature of Time Warner’s investment here make it very different from Google’s gigabit build out in Kansas City. As a final quibble, I’d also note that Google is building out its network in a way designed to challenge the economics and status quo associated with residential broadband, while Time Warner is merely continuing a gradual investment in its business.

Comparing TWC’s investments in its network to Google’s network investments, just because TWC tosses the words gigabit and fiber around, is like comparing McDonald’s oatmeal to the porridge nutritionists recommend because both contain oats. Don’t buy into that hype.

Time Warner photo courtesy of (CC BY 2.0) Flickr user The Consumerist. NYC skyline photo provided by Shutterstock.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Who and what to watch in the new era of the living room
• A look back at mobile in Q1
• Netflix may suffer from limited mobility

XO sells transport and intra-city capacity to enterprises, cable companies, wireless and wireline carriers, and internet companies looking to move massive amounts of traffic between point A and point B. For the upgrade, XO is using Nokia Siemens Networks’ latest-generation dense wavelength division multiplexing (DWDM) technology to boost the bandwidth carried by a single wavelength of light from 10 Gbps to 100 Gbps. NSN’s platform can support 96 simultaneous wavelengths so the theoretical total capacity of the network is 9.6 Terabits per second.

This isn’t the same thing as the fiber network Google is rolling out in Kansas City, which connects individual homes with high-capacity fiber strands. But fiber-to-the-home services like Google’s will need transport networks like XO’s to move the enormous sums of traffic they generate to their final destinations. In addition, the proliferation of streaming video, cloud computing services and just our general hunger for more data on more devices will feed the demand for souped-up long-haul links.

XO’s may be the first US operator to supercharge its inter-city fiber links with 100G but it certainly won’t be the last.

Photo courtesy of Flickr user Pasukaru76.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• How HR can make the case for workforce analytics
• The 2013 task management tools market
• How consumer media will change in 2013

“There’s no sense selling a product at a loss,” said Google CFO Patrick Pichette (just look at Google’s Nexus 7 tablet). “But it’s not only about profits, it’s about changing the access costs.” His goal and Google’s goal is to bring the same efficiencies that have helped create cheaper, smaller and more powerful computers and create a cost and improvement curve for broadband access that resembles the curves for compute storage, as the chart below illustrates.

And Google may have a found a way to do that — both in terms of constructing and operating a fiber to the home network — by using its engineering team, existing consumer technologies such as QR codes and social engineering to influence how users sign up for access. Existing ISPs should take note — what Google has done here has fundamentally lowered the cost of building and deploying a network. It was cagey about if and when it would take its fiber-to-the-home show on the road, but if it does, it will pummel existing ISPs on price and service, have repercussions throughout the carrier equipment industry and entice a lot of end consumers to take on a more active role in marketing Google’s broadband.

How Google cuts costs

Delivering broadband is a capital-intensive business, with Verizon spending $23 billion to spread its fiber to the home service to 17 million homes. Analysts estimate that it cost Verizon roughly $670 to run fiber past each home in its footprint. That cost varies depending on a huge number of factors, ranging from how far apart homes are to whether or not Verizon could string fiber from telephone poles rather than bury it. Google doesn’t give its costs, and so far vendors are mum, but here’s what we do know.

It makes its own gear: From the infrastructure on the back end to the TV and Wi-Fi routers in the home, Google has built its own stuff. Most carriers rely on outside vendors to sell them networking gear and even set-top boxes. However, like Iliad, the operator in France that provides the Free mobile and wireline network, Google has built its own equipment. Several sources have told me that Google has ordered fiber gear from companies such as Ciena, asked them how the boxes work and then sent the optical engineers on their way.

Kevin Lo, the general manager of the Google Fiber business, told me that from the time the Google Fiber project was announced in February 2010, engineers have been working on the gear. There are two advantages Google gains here. The first is that it’s not shelling out hundreds of thousands of dollars on specialty equipment built for ISPs, but rather taking the most basic elements of a network and assembling them into custom gear, much like it does on the data center side. The second is that it can control all of the physical infrastructure that its network relies on — updating and tweaking it as needed.

It uses social engineering: It’s accepted that one of the most costly elements of building out a fiber network is the physical labor associated with strong cable, digging trenches and hiring people to terminate the fiber into the home. Google has already strung cable on power lines throughout Kansas City and lowered those costs by working with the local utility and AT&T to get access to the utility poles without having to pay high fees.

But to reduce the cost of the actual last mile to users’ homes it’s telling people in Kansas City that if they want to be the first to get fiber, they’ll have to convince their neighbors to sign up. The goal is to get a critical mass of between 5 percent and 25 percent of the homes in a given neighborhood (Google calls it a fiberhood) committed to signing up for Google Fiber before ever sending out technicians. Residents have until Sept. 9 to get their fiberhood on the leaderboard before Google starts rolling out its fiber.

Google’s Milo Medin and a Google fiber product manager.

Milo Medin, the VP of access services at Google, explained that with this model the folks in the first fiberhood will have their access within a week. This is also why the free service is so important to Google. If people buy into that process, it can get homes attached in those initial bulk deployments and reduce the number of times Google has to send out trucks and technicians. Medin says the $300 initial connection fee will cover the costs associated with the deployments — it’s not doing that at a loss either.

It will use QR codes and the Google Play store to change your relationship with set-top boxes and routers : I’ve already covered the cost savings at the core network and the last mile access, but the final place Google is shaking things up is in the home. Customer premise equipment is the bane of the ISP industry. Those boxes are expensive so many cable providers and telcos rent them to users, which drives users nuts. Users also are slow to update the devices, which can limit the type of services ISPs can offer and in many cases force a technician to come out and install them.

Google has built its own hard drive to act as a DVR, a TV box to provide channels and a network box that acts as a modem and provides Wi-Fi connectivity in the home — cutting out traditional providers such as Arris, Scientific Atlanta (Cisco) and others that make such gear. Medin says that those boxes will have a QR code that a technician will scan. The box then sends its activation information to the cloud and the box is now provisioned and activated for that customer. Eventually consumers will be able to do this for themselves, perhaps after they order a box on Google’s Play store.

All of these things will help Google deliver a gigabit per second to the home at a profit. Granted, that profit might not be as large as the broadband profits that Comcast or AT&T currently enjoy, but it’s a profit. And hopefully regulators and average consumers will look at what Google is doing and ask themselves, “Why are the Comcasts and AT&Ts of the world complaining about how much it costs to serve up broadband when Google can deliver 100 times the traditional ISP’s top speeds for the same or a lower price.”

If they don’t ask, then let’s hope Google will continue its expansion. When I asked, Medin wasn’t direct, but said, “This is a beginning.”

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• 12 tech leaders’ resolutions for 2012
• Who and what to watch in the new era of the living room
• CES 2012: a recap and analysis

Unlike me, TWC has cash at its disposal and is offering $50 for information on Google’s fiber deployment, according to the poster a tipster sent me after a visit to the Time Warner offices in Kansas City. Which implies that Time Warner is concerned enough about the potential competitive threat that Google’s deployment represents to suss out all the details it can.

I can’t blame it. I’m pretty eager for tips on Google’s fiber deployment myself.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• A near-term outlook for big data
• Q4 Wrap-up: SOPA and the future of digital content
• Connected Consumer 2012: A year of consolidation and integration

New tech to cram more bits in your hertz.

These researchers have managed to achieve this epic speed at a distance of less than one meter using what researchers call “twisted signals.” According to Extreme Tech, which explains it so well:

These twisted signals use orbital angular momentum (OAM) to cram much more data into a single stream. In current state-of-the-art transmission protocols (WiFi, LTE, COFDM), we only modulate the spin angular momentum (SAM) of radio waves, not the OAM. If you picture the Earth, SAM is our planet spinning on its axis, while OAM is our movement around the Sun. Basically, the breakthrough here is that researchers have created a wireless network protocol that uses both OAM and SAM.

The resulting technology offers the spectral efficiency of 95.7 bits per hertz. To put that into perspective, today on Verizon’s LTE network, the equipment delivers 1.5 bits per hertz of spectrum. By delivering so much data per hertz of spectrum, the barriers toward building ever-faster networks as defined by Shannon’s Law would become fundamentally reset, allowing the next generation of engineers to build networks unimaginable to today’s generation.

Obviously, the usual caveats around new technologies apply. So far this is in the lab only. The speeds aren’t maintained for long distances and are based on lightwaves as opposed to radio waves (this means line of sight is essential). There’s no indication of how much power chips to deliver this type of speed would consume, and there’s no actual ecosystem in place to support or even twist those wavelengths.

That being said, there’s no Moore’s Law governing wireless networks, which is a real problem given how much data we are demanding via mobile networks. Granted, Wi-Fi and more efficient technologies will help, but we need a fundamental breakthrough on the physics side to keep up with wireless consumption. Terahertz spectrum and chips are one way, and these twisted signals might be another. As a plus, they could work on fiber optic networks too, which means we might see another boost in broadband capacity along our long haul and core networks.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• LTE-Advanced: what it is and isn’t
• LTE changes everything; LTE changes nothing
• Updated: Forecast: global mobile subscribers, 2010-2015