The clouds are gathering, and IBM couldn't be happier.

In conjunction with the Georgia Institute of Technology and Ohio State University, Big Blue recently announced that it will create the Critical Enterprise Cloud Computing Services facility, a "prototype computing cloud that links data centers from the two institutions," according to an IBM statement.

The recent profusion of the term "cloud computing" has left some people wondering what's behind the marketing vapor. In the context of IBM's work, cloud computing is a virtualized data center with automated management of functions that once required hands-on attention from IT staff.

Through its joint efforts with researchers and students at the two schools, IBM is pursuing a twofold goal: using the university brain trust to identify and begin developing computing technology that it can eventually commercialize, and seeding the next generation of IT engineers with the skills needed to manipulate technology that baffles ordinary computer users.

For those new to cloud computing, Frank Gillett of Forrester Research suggests a review of data centers and virtualization.

Before the emergence of virtualization - and still today - IT departments often loaded one application on one server, even if that meant utilizing just 40% of the server's capacity, because they wanted to keep the data isolated and guarantee reliable application performance. Virtual servers maintain that isolation while drastically increasing utilization rates.

"The whole virtualization thing was kicked off by the idea of subdividing a computer - having one computer pretend to be many in order to put more things on a computer that otherwise was not very busy," Gillett said.

IBM and other providers working in the cloud apply the virtualization concept across vast data centers. "When we talk about cloud computing, what we're actually talking about is scaling that up to very large numbers of machines supporting very large numbers of transactions," said Matt Ellis, IBM's vice president of autonomic computing.

In the broader business and consumer worlds, cloud computing is the engine that drives many popular Web services, including Google searches and the software-as-a-service applications that vendors, from SAP to Microsoft, now deliver over the Internet.

Anticipating that the model will only gain traction in the coming years, IBM wants to position itself as a cornerstone provider. The company is particularly interested in the automation of these vast and virtualized data centers. When dealing with thousands of servers and hundreds of thousands of virtual servers, Ellis said, "automation becomes a must-have. It goes beyond the human capacity to be able to deal with the level of change and the rate of change."