Blog Post

GoogleNET Going Global

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

More than 15 months ago, Google (GOOG) CEO Eric Schmidt publicly admitted that he wanted to build an infrastructure for what would be a global $100 billion company. Given that the search giant’s revenues are projected to reach just over $30 billion by the end of this decade, the claim was audacious — but not one to be taken lightly.

In order to build this leviathan enterprise, Google will have to continue to build its back-end systems at a rapid clip, but more importantly, it will also have to keep investing a copious of money to control the oxygen of the New Net Economy: bandwidth.

I wrote about the GoogleNET back in 2005 for the now-defunct Business 2.0 magazine and have been following this pretty closely, though not writing about as often.

The latest news from the telecom world is that Google is now looking to get into the transoceanic cable business. Transoceanic cables sit at the bottom of the sea, using capacities that run into the terabits per second to connect countries around the world.

There are rumors out there, first reported by Communications Day, that Google is heading up a new trans-Pacific undersea cable called Unity. (I guess after going to China, they can’t call it Do-No-Evil-1)

Communications Day understands that Unity would see Google join with other carriers to build a new multi-terabit cable. Google would get access to a fibre pair at build cost handing it a tremendous cost advantage over rivals such as MSN and Yahoo, and also potentially enabling it to peer with Asia ISPs behind their international gateways – considerably improving the affordability of Internet services across Asia Pacific.

Alan Mauldin, who tracks the sector for research firm TeleGeography, thinks that these are early discussions that might not turn into anything tangible. But regardless of the outcome, he notes that the talks themselves point to Google’s Godzilla-like appetite for bandwidth.

Trans-Pacific routes are among the busiest and most expensive in the world. The difference in the pricing between trans-Pacific and trans-Atlantic routes is akin to the difference between Bloomingdale’s and The Gap. Thanks to the broadband boom in Asia and growing economic trade between Asia and the U.S., the demand on trans-Pacific routes has exploded: between mid-2006 and mid-2007, Internet traffic in these routes was up 41 percent, according to TeleGeography.

At present there is 3.3 Tbps of lit capacity, but it is expensive. Google has to pay a ton of money for this bandwidth, just like everybody else. Sure there are new cables being put into place, and others like Pacific Crossing-1 and the Japan-U.S. Cable system are being upgraded, but that doesn’t solve Google’s insatiable need for bandwidth.

The Mountain View, Calif.-based company is putting a lot of resources into India, China, and South Korea as part of its global expansion. It’s also betting big on video, and has grand plans for offering more and more web-based services. All of that requires bandwidth — a lot of it.

By becoming part of the consortium, Google gets two things: wholesale prices for its bandwidth needs, perhaps preferential treatment, and the ability to still get to share the cost and risks associated with owning a sub-sea cable. Even if this does get done, it will be a few years still before Google can get its hand on all that capacity.

16 Responses to “GoogleNET Going Global”

  1. Google is the new Bell System.


    From servers to switches to optical transport gear, a great new network is taking shape. I wouldn’t be surprised if at some point, bits from a google server through their proprietary switches to other google data centers and servers aren’t even standard TCP/IP anymore… GCP/IP anyone?

  2. Brent Hopkins

    I would like to see some forward-thinking initiative with some muscle behind it, working on true wireless networking. Backhaul is currently the foundation of wireless communication, but need it be? The expense of it is astronomical and ultimately will require too many resources to maintain. I believe much of the current ‘forward thinking’ is actually backward thinking with concepts mired in an outlook dating back to the telegraph. The FCC and its overseas counterparts are smothering innovation by treating ‘radio’ spectrum as a scarce resource. In fact it is a vastly underused resource with deliberately created ‘scarcity.’ Let the mighty brains of Google chew on that: ultimately it would be a more efficient allocation of resources to go truly wireless.

  3. Frank A. Coluccio

    The French, too, are said to be planning a transpacific route, as reported by CommsDay, which is the same publication that leaked the Google story. Hm..

    This leaves room for further speculation about the original two speculations, since the carriers that are said to be partnering with Google haven’t been identified (rumored?) yet, either.

    In any event, I posted this story last month to another forum, along with a curious tidbit concerning Bangladesh’s submarine “miscreant” woes:

    8/28/2007 | Submarine Cables:

    [1] France in the Pacific

    [2] Bangladesh’s “Miscreant” Woes


  4. Why not be a telco indeed.. with the Google Mobile OS on the way, and with Google’s connections to the iPhone.. why not use the infrastructure and offer large enterprises full VOIP over Google infrastructure.

    Google Phone, running a Google OS, powered by Google services running over a Google infrastructure for a subscriber like service.

    Consumer skew gives Google another outlet for it’s advertising money machine, Enterprise outlet gives a high-revenue entrenched business.. and is yet another kick to the already tender regions of Microsoft’s flailing Live platform..

  5. A wireless network is feasible given this development. Backhaul and backbone would be available to just build on top. Why not be a telephone operator with all this infrastructure.

  6. Ragunath Padmanabhan

    I think at some point in this ambitious growth Google will start realizing that it needs to operate like a collection of small entities – with responsibilities and ownership equitably shared. And what will remain Google about these small entities are shared resources, best practices and above all a set of guiding principles. Dee Hock did this with Visa. What would be even better is the synergies between with I’d love to see within my lifetime (I’m 35) Google becoming a word that simply refers to a certain way of organizing economic and social activities such that there is no evil on the economics side and unlimited good on the social side. What else is there to dream about when you are already in the top ten list of billionaires?

  7. Once Google has all this bandwidth, will their web services put just about everyone else out of business? For example, will Google replace investment banks like Goldman Sachs and Morgan Stanley? Will Google buy News Corp from Rupert Murdoch and own the Wall St. Journal? Will Google buy Walt Disney (and all its assets) and just turn Disney into a giant YouTube? Google is already starting to look scary as if they truly could rule the world (and make Microsoft’s once monopoly look like child’s play). What would Larry Lessig say?