& cplSiteName &

AT&T Looks to Push Industry Faster to 400G

Carol Wilson
10/28/2016

AT&T's announcement Friday of plans to test 400 Gigabit Ethernet data speeds in 2017 puts the network operator at the forefront of the industry in that effort, according to Heavy Reading Senior Analyst Sterling Perrin.

It also reinforces the company's intention to move away from specialized optical routers toward more commodity hardware. (See AT&T to Test 400G in 2017.)

AT&T Inc. (NYSE: T) said this morning it will launch a three-phased trial next year, starting with a test of its global backbone between New York City and Washington, using Coriant optical gear (a major surprise there), continuing with a test of a single 400Gbit/s wavelength in a metro network using Ciena Corp. (NYSE: CIEN) gear and concluding with the first test of a 400GbE open router platform, which uses commodity hardware and puts as many functions as possible into software.

The latter is part of AT&T's aggressive push away from purpose-built hardware, as part of its adoption of SDN, as outlined last month by Tom Anschutz, distinguished member of technical staff, at Light Reading's NFV & Carrier SDN event. (See AT&T Ready for Next-Gen NFV.)

"This is all very forward looking," Perrin says. "AT&T is trying to be ahead of the technology curve. They are definitely out in front on this."

AT&T is determined to push the industry ahead as quickly as possible, to prepare its networks for the hyper-connected world to come, says Dan Blemings, director of Ethernet product management for AT&T Mobile and Business Solutions.

"We are at least several years away from putting a product in the market, but we know that day is coming," Blemings tells Light Reading in an interview. "Today represents a very proud moment -- we are here to say we are stepping out in front; we are going to move the industry forward as fast as we can because this is critical to our future."


Want to know more about 400-Gig developments? Check out our dedicated 400G/terabit section in the optical network channel here on Light Reading.


The first phase, testing 400Gbit/s speeds across AT&T's long-distance network, involves using two 200Gbit/s wavelengths, as opposed to a single 400G wavelength, because to date there isn't a 400-Gig transport option. "There is the interface and there is the transport -- they are two different things," analyst Perrin says. "There is still a way to go on 400-Gig transport."

He is surprised the test is happening on Coriant gear, since the optical network vendor has been struggling of late. "They don't seem to have the business coming in to create the investment necessary, and they are being perceived by their competitors as weak -- their competition is going after their accounts pretty heavily," he says

Blemings says, however, that AT&T has worked with Coriant in the past and found them to be a good partner and likes the technology they bring to this trial.

"We use some Coriant in our network today. They are a great innovator, and they were a great partner to work with on the trial," he says. "They have some interesting hardware that is interesting for us to move forward with, so it made a lot of sense. We have a relationship with them now and they have awesome gear that we wanted to trial, so they were a natural fit for this."

The metro trial, which is phase two in AT&T's announcement, does specify a single 400Gbit/s wavelength and that makes sense for connecting large data centers within the metro space, Perrin says, given that traffic within the data centers is moving rapidly from 10 Gbit/s to 25 Gbit/s to 100 Gbit/s.

"Connectivity between data centers would go up," he says. "Otherwise you have a lot of stuff at the data center at 100 GigE being connected to another data center at 100 GigE, so [operators] do need to plan for ultimately connecting data centers at higher speeds."

Testing AT&T's plan for a disaggregated router at the 400-Gig level also makes sense, given that the network operator already has functioning 100-GigE routers today. Investing in the process of disaggregating hardware and software for that technology doesn't have much, if any, of a payoff.

"If you look at 400 GigE, it's an opportunity to do something new," Perrin says. "They know they have a long window of time before it becomes a mainstream application, and when they have figured it out, they hope the time is right and they've discovered a better way of doing things."

Blemings describes an open platform router as critical to AT&T's future, as it uses virtualization to create a more flexible network which can deliver massive amounts of bandwidth needed going forward.

"Four-hundred gigabits is definitely our future especially as we move into a hyper-connected world," he says. "Having a disaggregated router or this open platform at 400-Gig is going to be critical -- that is how we are moving as a company. Having hardware that uses merchant silicon and non-proprietary pieces of software is what will enable us to innovate faster. We make that hardware responsive to the demands of our network in real time. That is a big part of what Open ROADM is about and what the disaggregated router is about. It allows AT&T to customize the equipment, so we can configure it to the needs of AT&T."

— Carol Wilson, Editor-at-Large, Light Reading

(4)  | 
Comment  | 
Print  | 
Newest First  |  Oldest First  |  Threaded View        ADD A COMMENT
Sterling Perrin
Sterling Perrin
11/1/2016 | 9:20:16 AM
Re: I don't understand the surprise
The comment sounds subjective but the surprise is based on objective data: North American long haul DWDM has become incredibly concentrated - Ciena and Infinera have ~70% market share, with Nokia (ALU) a distant third. If a tier 1 long haul win is not one of these three suppliers today, then it's a surprise.

The North American metro optical market (DWDM/packet-optical) is more fragmented with a number of suppliers in play. However, Coriant has been losing share in NA and is currently at just 10% share. So, a Coriant win with a tier 1 operator, given its share, is a also a surprise.

Sterling
danielcawrey
danielcawrey
10/31/2016 | 8:16:27 PM
Re: I don't understand the surprise
I'm all for more speed, and given the way we use the internet today, we're all going to need it. Video. VR. Whatever else – all at a constant rate – is going to burn through the speeds we have today, that's for darn sure. 
Carol Wilson
Carol Wilson
10/31/2016 | 12:32:07 PM
Re: I don't understand the surprise
Check out Sterling Perrin's comments - he explains the reason he was surprised by the choice of Coriant. It's not a comment on the technology so much as the company itself. 

Sterling actually spoke highly of the new DCI product launched last year. But as he notes, the company has gone through layoffs (I know, who hasn't?) and competitors to Coriant have been going after its customers, thinking the company is at a weak point. 

The AT&T announcement could be enough to change that thinking - generally big players like AT&T don't choose vendors when there is a lot of uncertainty.
Projectl28482
Projectl28482
10/30/2016 | 12:50:26 AM
I don't understand the surprise
Carol,

why did you write "a major surprise there" concerning Coriant and its selection for the AT&T trial?

Coriant has now the best solution for 200G and 2x200G with respect to the competitors.

As easy as it is...Things will change soon.
Featured Video
Upcoming Live Events
October 22, 2019, Los Angeles, CA
November 5, 2019, London, England
November 7, 2019, London, UK
November 14, 2019, Maritim Hotel, Berlin
December 3-5, 2019, Vienna, Austria
December 3, 2019, New York, New York
March 16-18, 2020, Embassy Suites, Denver, Colorado
May 18-20, 2020, Irving Convention Center, Dallas, TX
All Upcoming Live Events