Around 18 carriers now connect to CENX, says Sandy Brown,
the new CEO of the first carrier ethernet exchange. The
technology will dominate corporate services within a few
Sandy Brown: former AT&T executive has taken over as CEO of
CENX, the world’s first carrier-neutral carrier
CENX, the operator of the first carrier ethernet exchange, has
hired former AT&T executive Sandy Brown as its first CEO.
The decision marks the next stage in the development of a
project that aims to create a business that some believe will
be worth $40 billion by 2013, as carriers and their enterprise
customers switch to ethernet technology for their long distance
and international links.
"It’s a pretty exciting time," says Brown, whose
appointment was announced in early July. "We’re at
the beginning of a pretty significant transformation of the
CENX — Brown and others in the company pronounce it
"see-nix" — was created by a group of evangelists for
carrier ethernet, notably Nan Chen, president of the Metro
Ethernet Forum, which is advised by the venerable Bob Metcalfe,
inventor of ethernet back in 1973.
It has taken most of the subsequent 37 years to get ethernet
out of the office and under the streets, into metropolitan
networks; and the next stage — inter-city and
international networks — is starting to emerge now.
"Ethernet is generally cheaper, more cost effective in
equipment terms and cheaper in cost per bit transmitted," says
Brown, days after his appointment. "And it has better scaling."
You can go from 10 megabits a second to one gigabit a second
"with typically no hardware changes", he says.
More people are familiar with ethernet technology than with
traditional telecoms protocols, he adds: remember, it is a
networking technology that has been in virtually every office
in the world for three decades. Every in-house IT department
has people who understand ethernet and work with it daily.
"Ethernet is a better solution for businesses as they grow," he
Chen, co-founder and president of CENX, welcomed
Brown’s appointment: "Sandy brings extraordinary
operational experience and a shared vision of the profound
changes and challenges carriers face as ethernet rapidly
becomes the dominant technology for service delivery." Chen was
a speaker at Global Telecoms Business's Innovation Summit in
So what is CENX? It is an imaginative bid to do in the ethernet
telecoms world what has been standard in traditional TDM
telecoms almost since the industry started. No one wonders that
a call from Anne to Bob may travel through the networks of two,
three or four telecoms operators, depending on where Anne is
and where Bob is.
Matching networks with customers
Ethernet operators are focusing on winning business from large
corporate customers. "Our role is recognising a basic reality,"
says Brown. "Carriers want to provision customers on their own
networks, but customers’ needs and
carriers’ networks don’t always
Because of this, "carriers need a good way of buying offnet
services". If a carrier can provide ethernet connections to
nine out of 10 of a customer’s locations, it needs
to recruit a competitor to provide the tenth link.
But how does it connect? That’s the role of CENX:
it is building a series of carrier-neutral ethernet exchanges,
places to which a number of competing carriers will connect.
"The exchange makes it easier for customers to connect," says
"Customers" in this sense are operators, the approximately 18
carriers — Brown is guarded about giving the exact
number — that have signed up and connected into CENX.
The company itself "doesn’t make any
announcements" about members, said an official.
However it is known that Verizon is connected: the
carrier’s Quintin Lew and CENX’s Chen
were presented with a Global Telecoms Business innovation award
in June 2010 in London for their work on the project.
Interxion, which provides co-location data centre services in
Europe, announced in April 2010 that it would connect from
June. RCN Metro Optical Networks, a US carrier, has also
announced that it has become a member, as have Optimum
Lightpath — a metropolitan New York carrier —
as well as Level 3, XO and others.
Ethernet service translator
Each connects through a one-gigabit or 10-gigabit pipe, and the
CENX equipment has a service translator, that does its best to
map carrier A’s offerings to those that are
available from carrier B. This is a complex process: "A few
things are a little problematic," admits Brown.
The relationship between any two carriers is negotiated between
those companies: CENX does not arrange to business, just
facilitates it in its equipment. "CENX is a marketplace. It
allows the business to work more effectively," he says.
End-users, the multinational corporate companies, do not have
any direct relationship with CENX.
The company is also developing a set of online tools for
operator members, including a system that allows them, if they
want, to put details of their network inventory online. So
operator A, looking for a network that will deliver that tenth
location to its corporate customers, can search through likely
candidates before talking to them, and then arranging an
interconnection via CENX with the one offering the best deal.
"We don’t get in the middle of the business
transaction," says Brown. It’s up to member
companies to decide what locations to offer, if any, via the
online system, and they can show the different service types
that are available.
They can offer the data in "pretty much any form", he adds
— via a list of geographical areas, a list of
administrative areas, a set of polygons on a map, or a file
listing buildings or zip codes or postal codes served. "It
depends on the strategy of each carrier. Some are more building
oriented, and others are more geographically oriented."
CENX will earn its revenue from three sources, he adds: it
charges for each initial connection; there is a service cost
for connections; and there is a speed-related cost. The
inventory service "is part of the deal".
London and Hong Kong
CENX started in the US with hubs in New York, Los Angeles and
Chicago, and it has opened its first hub in Europe, in London.
"Hong Kong will probably be next," says Brown. "And we will
have one in Miami for South America."
Now, the company has "a very aggressive build-out plan to grown
from six to 20 locations by the end of 2011", he says. "We are
driven by our carrier customers." CENX is aiming at "the major
European telecoms areas and the Asia Pacific region".
So who funds CENX? "We don’t talk about the
investors," says Brown. "It’s backed by some
really extraordinary folks."
Chen is clearly one. He has a track record of start-ups,
including Atrica, which Nokia Siemens Networks bought at the
end of 2007. But for any detail, the CENX website has nothing
to say and Brown has little more: "These are not dumb money
investors," he says. "They have a vision about where CENX could
be. It’s very much a shared vision."
CENX and Verizon’s submission to the 2010 GTB
Innovation Awards gave some background. "CENX, Verizon and
other service providers developed the carrier ethernet neutral
exchange model for two years. The first carrier ethernet
exchanges became operational in October 2009 in New York, Los
Angeles and Chicago," it said. CENX and Verizon started to work
together in 2008 but their partnership was not announced until
The submission added: "CENX and Verizon also worked closely
with several large service providers to create an innovative
approach to provided scalable global growth for carrier
Says Brown: "Carriers have been interconnecting for ages, but
they used different approaches. What changed was when ethernet
crossed a couple of thresholds. Carriers are starting to buy
and sell ethernet at some scale."
And so there is a need for CENX to interconnect services in a
carrier-neutral location. CENX does not provide its own
long-distance services, he emphasises: it does not compete with
the companies that connect to it.
Having started at the start, CENX can only get bigger as
ethernet grows, thinks Brown. "Over the next five years the
vast majority of new circuits will be ethernet access end to
end," he says. Vast majority? "Yes, 80-85%," he adds.
Analysts certainly agree with the potential for carrier
ethernet revenue, suggesting it will be nearing $40 billion in
only three years.
Access to capital
So how much has CENX available in resources? Just as with the
related question about investors, Brown did not want to give
detailed answers. "We are working with very significant
investors," he says. "Access to capital is not a challenge. As
we grow, that will enable us to grow."
CENX is still a small operation, with around 30 full-time
staff. It uses other companies’ locations. "We are
not in the real-estate or co-location business. Our investments
are in systems and tools."
He is unwilling to say whose kit the company uses: "Our
switches are best of breed, industry grade products", he says.
"Our partners need to be assured that it’s carrier
In late July 2010 Canadian OSS company Exfo announced that it
had an agreement to supply equipment to CENX to monitor service
level agreements of offnet connections — the sort of
thing companies connected to CENX will need in order to
reassure their end-user corporate customers.
"We’re building our own system processes," says
Brown. Partners "need to be assured that it’s
carrier grade. Carriers have to be able to certify the services
that they offer to others."
Brown himself comes from a solid telecoms background, having
worked for Bell Labs — when it was part of classic
AT&T — and then AT&T itself for most of his
career. "I worked on PBXs at Bell Labs a long time ago," he
says. He spent the first half of his career working with
equipment, including PBXs and fax machines, and the second on
services — "predominantly data networking, including
getting AT&T into the internet". His last AT&T
assignment was as a vice president, responsible for
multi-billion dollar lines of business including its carrier
But his conversion to the ethernet cause "began a long time
ago", he recalls: back when he was "in PBX world and local area
networks were just getting started. I realised ethernet was
going to be an alternative way of communicating. It was clear
then that it was a long way out, but you could see that it was
Since then, ethernet has progressed from being a technology for
local area networks, for which it was primarily designed by
Metcalfe and his colleagues. At the time, back in the late
1970s and early 1980s, there was great controversy about which
of half a dozen local area network technologies would win. "We
know who the winner is now," says Brown. GTB
Nan Chen of the Metro Ethernet Forum and co-founder of CENX at
the GTB Innovation Summit, June 2010:
First GTB report about CENX: