First Google, Now IBM, Cast Cloud Computing in Carolina

By Clint Boulton  |  Posted 2008-08-01 Print this article Print

UPDATE: The story was corrected to reflect that the data center is in Durham County, N.C.

IBM is pumping $360 million into a cloud computing data center in Durham County, N.C., just a few months after Google opened a $600 million data center in Lenoir, N.C.

Cloud computing is the catch phrase given to computing resources piped over the Internet, but hosted by a provider such as Google, IBM, Yahoo or Amazon.

HP, Intel and Yahoo just launched a global cloud computing effort for academia earlier this week, so the space is certainly abuzz. The cloud arms race is on.

As for IBM's new facility, what IBM will put in its data center boasts similarities and differences to what Google has done in Lenoir. Let's start with the hardware.

Google likes to run an amalgamation of standard hardware in its data centers and is pretty secretive about what exactly it stacks its racks with.

IBM, a server maker, will run x86- and Power-based servers along with IBM's System z mainframe and other IBM hardware, storage products and management software, such as Tivoli. I expect IBM will employ several iDataplex machines, which are basically blade servers that IBM has dressed up as cloud servers.

The software running in these data centers is what Google and IBM have most in common. Last October, Google and IBM embarked on their pact to help drive cloud computing in universities.

For these data centers, Google and IBM run various flavors of the Linux operating system, Xen systems virtualization and Apache's Hadoop project, an open-source version of Google's MapReduce distributed computing software, and the Google File System.

Rich Lechner, vice president of Cloud Computing Strategy for IBM, told eWEEK's Scott Ferguson that IBM would use some of the distributed computing software above, probably Xen, Linux and Hadoop, as well as some of IBM's own software. 

How Google and IBM are using the Internet as a computing platform is where the companies diverge. Google essentially powers itself, or the search service and applications it provides to customers.

We as consumers all access these from PCs, laptops or smart phones. We do a search query and it races over the wires to get a response from one of Google's many data centers, which are processing data in parallel all over the world.

Google does power some enterprise apps, including security SAAS, enterprise search and its App Engine hosted apps infrastructure, but the company's bread-and-butter is serving general search to the world.

IBM will expressly be providing computing and storage resources that companies of all sizes can procure to run their own businesses. Imagine small-to medium-sized businesses with no technology or IT budgets as IBM's customer base.   

In May, I had the opportunity to attend the groundbreaking ceremony for the Google data center that it set up in Lenoir.

It was a momentous occasion for the town. The excitement was palpable even though no one was permitted to enter the data center. How fascinating is a football field of standard rack-mount servers anyway?

No, the real hype was all the jobs the data center had created for Lenoir, a town that needed it.

And now, IBM has brought that vibe to Durham County.

IBM and Google, for all of the similarities and differences in their respective approaches, are casting quite the cloud in North Carolina. |

Submit a Comment

Loading Comments...
Manage your Newsletters: Login   Register My Newsletters

Rocket Fuel