A flurry of new projects from IBM signal the company's slow but steady shift to cloud computing. IBM is pitching small-scale use cases and "cloud wins" to demonstrate where it sees the opportunity today for cloud: infrastructure automation and virtual desktop services. It also offers public cloud services, but customer numbers remain small.
For instance, IBM is helping the University of Bira in Italy deliver mapping and market data via
In reality, no IT department on earth can afford to invest in cloud without knowing what it is going to cost them.
Dennis Drogseth, vice president at Enterprise Management Associates
In Malaysia, IBM is helping state-run Multimedia Development Corporation run an animation cloud that can be accessed over the Internet to render computer-animated images. In addition, last week, IBM launched a second cloud services data center in the EU, and it's heading an EU research effort named "PINCETTE" to improve the reliability of complex information systems. None of these are enterprise-sized deployments.
"The pickings are pretty slim, but literally everything anyone is doing [in the cloud] is pretty innovative, so they want to push that," said Sean Hackett, research director for the 451 group. Hackett said that IBM sees plenty of room in the short term for research and experimentation, as evidenced by its university projects and new research groups.
IBM does offer a broad portfolio of products and services it calls cloud, which include online services, "cloud-in-a-box" appliances and data analytics software. But for now, the company cheerfully admits that large-scale total IT transformations into cloud environments are far off.
Bruce Otte, senior marketing manager for cloud at IBM, said that cloud computing represents a fundamental market and technology shift, but added that IBM is looking to smaller, bite-sized cloud projects first. He said academic organizations represent the perfect place to experiment, and bright-eyed students and idealistic administrators are a plus, too.
"After all," he said, "for lack of a better term, the labor is really cheap!"
IBM's long-term cloud goals
In the long term, Hackett said IBM wants to become a "cloud hub," a central provider of services that IT departments can tap for highly standardized and integrated services for short-term, on-demand use as well as providing infrastructure where appropriate. It sees itself as eventually able to provide instant access to complex and highly automated data services that enterprises can use with existing backend systems and data without much work.
"It's a solutions approach; it's pre-integrated," he said.
Hackett said IBM has spent $8 billion over the last couple of years building out a dozen data centers around the world, a significant capital investment even for a company of IBM's size. Those data centers will help power IBM's services arm as the world increasingly and gradually turns to cloud computing. But Hackett said IBM would not be getting into the IT Infrastructure as a Service (IaaS) business, much like Amazon Web Services, as its bread is not buttered in just provisioning capacity.
"They've made a ton of money on things being complicated," he said.
Analyzing IBM's cloud choices
"I think IBM recognizes the responsible way to approach a cloud direction," said Dennis Drogseth, vice president at analyst firm Enterprise Management Associates. Drogseth said IBM is working in small steps to streamline its existing technologies and make them applicable to wider parts of the business.
He pointed to IBM's October 2009 release of VMControl, automated virtualization control software that is tightly bound with IBM's Tivoli software. "Process automation is really key," he said.
Drogseth noted that IBM has a huge investment in cross-platform tools already. It has a large portfolio it can apply in service automation and service accounting, and understanding how the money stacks up is fundamentally important within a large enterprise.
"In reality, no IT department on earth can afford to invest in cloud without knowing what it is going to cost them," Drogseth said. "If they don't, it'll come around and bite them."
He said he believes IBM's current cloud computing offerings, like the Test and Development Cloud, online storage services and its recent acquisition of data integration vendor Cast Iron, point to a careful strategy of covering industry verticals and making sure IBM remains central to the enterprise as businesses become increasingly entangled in a web of online services.
"They certainly don't want to run up against a brick wall in helping verticals evolve simply because they weren't there in cloud," Drogseth said.
However, IBM also knows that simply being there to provide capacity is a dead end. Drogseth said that many are getting caught up in the current cloud hype, and while on-demand access and pay-as-you-go, elastic hosting is neat, infrastructure vendors sometimes forget that infrastructure isn't the endpoint; it's less important than actually delivering IT services and accounting for them.
Drogseth and other analysts are relatively bullish on IBM's strategy, pointing out that the firm has a very long history of making money, a customer base that includes all of the world's largest enterprises, and no shortage of innovative ideas or resources.
IBM risks losing a march here or there to smaller competitors, and it's never, ever going to be cheap, but unless it fails badly on reliability or service delivery (IBM's mainstays), it will find itself comfortably ensconced in the middle of the enterprise as the gradual shift to highly virtualized, highly automated, online systems comes to pass.
Carl Brooks is the Technology Writer for SearchCloudComputing.com. Contact him at firstname.lastname@example.org.