Google (Nasdaq: GOOG) is building a network so massive that several service provider specialists believe it could end up with one of the world's largest core transport networks, effectively building its own private Internet.
”This is huge,” says Hunter Newby, chief strategy officer with carrier connection specialist Telx, pointing to several recent indications of the Google network’s scale. “It’s scary. They’re not fooling around.”
Light Reading, which has previously reported on Google's telecom aspirations, has learned that Google is well underway at putting the pieces together (see Headcount: Great Googly Moogly!, Google Backs Powerline Carrier, and Google Talks the Talk). It is accumulating hundreds of thousands of square feet of carrier hotel space that could host giant server farms, buying up fiber, and issuing large RFPs for DWDM and Ethernet-based telecom equipment that could total in the hundreds of millions of dollars, according to multiple sources at carriers and equipment vendors.
By building a core of transport techologies and peering directly to the world's leading incumbent telcos and PTTs, Google could end up securing and controlling distribution of much of the world's Internet traffic, say the sources. Its massive server farms would have a direct link to the backbone.
Google did not return phone calls for this article. In the past, when asked about its telecom networks, the company has declined comment.
One indication of the size of the product is Google’s recent real estate activities. Crains New York Business reported yesterday that Google acquired the rights to 270,000 square feet of space at 111 8th avenue in Manhattan, which is a large telecom interconnection facility. That space is expected to be populated with telecom equipment and server farms running Google's search and other applications such as GoogleTalk (see Google Talks the Talk).
Another service provider operator, speaking under condition of anonymity, says Google's got big plans on the West Coast, where it is negotiating for large amounts of carrier hotel space and hopes to connect to Asia through the largest fiber peering points.
”I could tell you more but they would kill me,” says the interconnection specialist.
Why does Google want to do this? One idea is simply to reduce its telecom costs and peering fees, which many believe are significant. Another idea is that by building its own core network and focusing on Layer 1 and Layer 2, Google could control the distribution and security of much of the content and traffic distributed over the Internet. In a sense, it would be a higher-performance Internet, such as the research network, Internet2.
”My understanding is they want to do remote peering and transit bypass,” says Bill St. Arnaud, senior director of advanced networks at Canarie Inc., who has heard scuttlebutt about the Google network. “By building their own distribution network they don’t have to pay peering costs. Remote peering and transit costs are significant for all the big Internet players. So everybody is thinking of doing this.”
Naturally, equipment provider mouths are watering over the project. Several sources tell Light Reading that Google has issued a DWDM RFP that could be one of the largest Tier 1 service provider contracts, with the total value expected to be in the $100 million ballpark. Siemens AG, sources say, is particularly excited about bidding on the RFP, given its recent victory in similar DWDM networks at AT&T Corp. and MCI Inc., says one source. Others believed to be in the bidding include Ciena Corp. (Nasdaq: CIEN), Lucent Technologies Inc., Nortel Networks Ltd., and Huawei Technologies Co. Ltd.
Because Google is acquiring its own fiber and building Layer 1 and Layer 2 equipment at global interconnection facilities, it can create its peering points at large interconnection facilities, sources say. This would allow it to peer with global PTTs and keep much of its own traffic in a private Layer 2 network, securing and speeding up the performance of much of its traffic.
”This could create a shift in where public IP interconnects,” speculates Newby. “Traditionally, people went to Internet peering points. But because Google is so large, it could be the Internet. People would go there and never leave.”
if you are looking for something different and better than Netmeeting. look at vidrev, it is at vdrv.com try the download from the webloader during the trial. if you want this product for your company it can be free peer to peer, i would like anyone's imput on this. so let me know. regards noreen
The figures I gave were pretty accurate ($100M for a nationwide LH DWDM trasport system with lots of channels lit, and about $20M for the LH fiber of about 20,000km).
If you are suggesting you can't replicate a general-purpose, go-everywhere, try-to-be-everything-to-everyone style of network ala AT&T, MCI, Sprint, Level3, etc, you are correct.
I don't believe Google is trying to replicate such disasters. The original question only asked how much DWDM gear you could buy for $100M. If anything, Google would be trying to build a very lean purpose-built network for something, thus, would not require the massive infrastructures that the big guys have. In that context, $100M gets you a really big network.
Where is Honestly? I don't see him on this wagen, so i'm going to start the rumor he would hate to hear. Who can confirm the google crs-1 deal is on/done?
I would like to come back to the original question
If you look at the GOOGs WiFi secure solution, this is nothing more than a VPN tunnel from the user directly to the Google servers.
Look at Googles FAQ :
"If you choose to use Google Secure Access, your internet traffic will be encrypted and sent through Google's servers to the Internet. The data that is received will then be encrypted and sent back through our servers to your computer."
Does that mean that GOOG is going to let all internet traffic of these users pass trough their VPN servers ? Suppose this secure WiFi solution becomes popular, the google servers will see a HUGE amount of traffic passing. Could that be a reason to build an own, huge network ? Are we tomorrow sending all of our internet traffic first to Google before it is passed from there to the destination server of my preferred site ? This would really put GOOG in the driver seat. With one move, GOOG reduces all operators to bitpipe providers (from the end-user the Google network)
"The reason VoD services haven't taken off yet is because the content available isn't diverse enough, it isn't easily accessible, and it isn't offered at the right price."
I have to agree with this. Putting on my "consumer" hat, I have to say that I usually don't want any of the whopping 5 movies that are available on demand from my Cable company at any one time, and I don't want to pay movie theater prices for it. And the whole "mobile video" concept is stupid -- not because I wouldn't want to see a movie while traveling, but because video is much less appealing on a 2 inch cellphone screen or on a 100kb link.
What IS appealing is a video that is of broadcast quality and large enough to fill my laptop monitor. We are just on the cusp of this sort of service becoming available. BitTorrent is just the precursor, while we are waiting for the delivery infrastructure to support this in streaming form. This quality of video takes ~2Mb to deliver with today's best compression technology. 2Mb data services to the home have only been available for ~3 years and are only now becoming reliable & common enough for this to be feasable.
Based on this, and having kept my ear to the ground, I think the "video revolution" that people have been promising me for the past 20 years is actually going to happen. (I remember being in the ATM switch business in the early '90s, and someone asked "can I really get video on it?" to which, my cynical response was "yeah, IF you've got something to connect it to, and IF you've got some content to feed into it, and IF you have some applications at either end to set it up....". My cynicism is fading quickly.)
Perhaps that is because anyone other than GOOG would be missing a key piece of the revenue puzzle. It makes no sense to get the ball to the 99 yard line and drop it like a LVLT.
The blogs and comments are the opinions only of the writers and do not reflect the views of Light Reading. They are no substitute for your own research and should not be relied upon for trading or any other purpose.
To save this item to your list of favorite Light Reading content so you can find it later in your Profile page, click the "Save It" button next to the item.
If you found this interesting or useful, please use the links to the services below to share it with other readers. You will need a free account with each service to share an item via that service.