PBS | I, Cringely . November 17, 2005 - Google-Mart:
"The same follows for the rumor that Google, as a dark fiber buyer, will turn itself into some kind of super ISP. Won't happen. And WHY it won't happen is because ISPs are lousy businesses and building one as anything more than an experiment (as they are doing in San Francisco with wireless) would only hurt Google's earnings.
So why buy-up all that fiber, then?
The probable answer lies in one of Google's underground parking garages in Mountain View. There, in a secret area off-limits even to regular GoogleFolk, is a shipping container. But it isn't just any shipping container. This shipping container is a prototype data center. Google hired a pair of very bright industrial designers to figure out how to cram the greatest number of CPUs, the most storage, memory and power support into a 20- or 40-foot box. We're talking about 5000 Opteron processors and 3.5 petabytes of disk storage that can be dropped-off overnight by a tractor-trailer rig. The idea is to plant one of these puppies anywhere Google owns access to fiber, basically turning the entire Internet into a giant processing and storage grid.
While Google could put these containers anywhere, it makes the most sense to place them at Internet peering points, of which there are about 300 worldwide.
Two years ago Google had one data center. Today they are reported to have 64. Two years from now, they will have 300-plus. The advantage to having so many data centers goes beyond simple redundancy and fault tolerance. They get Google closer to users, reducing latency. They offer inter-datacenter communication and load-balancing using that no-longer-dark fiber Google owns. But most especially, they offer super-high bandwidth connections at all peering ISPs at little or no incremental cost to Google.
Where some other outfit might put a router, Google is putting an entire data center, and the results are profound. Take Internet TV as an example. Replicating that Victoria's Secret lingerie show that took down Broadcast.com years ago would be a non-event for Google. The video feed would be multicast over the private fiber network to 300 data centers, where it would be injected at gigabit speeds into each peering ISP. Viewers watching later would be reading from a locally cached copy. Yeah, but would it be Windows Media, Real, or QuickTime? It doesn't matter. To Google's local data center, bits are bits and the system is immune to protocols or codecs. For the first time, Internet TV will scale to the same level as broadcast and cable TV, yet still offer soemthing different for every viewer if they want it.
As for the coming AJAX Office and other productivity apps, they'll sit locally, too. Two or three hops away from every user, they'll also be completely backed-up by two to three data centers down the line. Your data never goes away unless you erase it. Your latency and system response are as low as they can possibly be made for a network app.
And remember the Google Web Accelerator that came and disappeared? It's back! Only this time the Web Accelerator will have the proper hardware and network infrastructure to make it worth using.
This is more than another Akamai or even an Akamai on steroids. This is a dynamically-driven, intelligent, thermonuclear Akamai with a dedicated back-channel and application-specific hardware.
There will be the Internet, and then there will be the Google Internet, superimposed on top. We'll use it without even knowing. The Google Internet will be faster, safer, and cheaper. With the advent of widespread GoogleBase (again a bit-schlepping app that can be used in a thousand ways -- most of them not even envisioned by Google) there's suddenly a new kind of marketplace for data with everything a transaction in the most literal sense as Google takes over the role of trusted third-party info-escrow agent for all world business. That's the goal.
All this is based, of course, on Google's proven network and hardware expertise. Have you seen Google's Search Appliance? They ship you a 1U prebuilt server. You connect it to your network, fill out a simple configuration screen, and it scans and indexes your web site (or sites) for you. Google monitors and manages it remotely, and sucks up the data and adds it to theirs. You just plug the thing in and turn it on. It just works. You need do nothing else to keep it running. Google understands how to do this stuff. Microsoft definitely does not.
And there lies the differences between the two companies. Last week, I wrote about Windows Live and Office Live as Microsoft's best attempts at pretending to be Google. And Google will do those kinds of applications, too. But they'll build them atop a network infrastructure that Microsoft can't match."