There have been no shortage of cloud computing announcements from IBM over the past 12 months, including a deal with Google to make cloud capacity available to university researchers. Last month, IBM announced Tivoli storage as a service and cloud management tools. At the time, Ric Telford, IBM VP of cloud services, told me that IBM has been offering on-demand computing to enterprise customers for nearly five years and that it has "several hundred" customer engagements involving its cloud services. (See "IBM Drifts Slowly Toward Mainstream Cloud Computing.")
IBM's cloud strategy is characterized by products and on-demand services that are aimed at IT departments and businesses, not at individual developers, entrepreneurs, and end users. While you can pay by the hour for Amazon Web Services with a credit card, IBM tends to require an annual contract for its services.
The United States Golf Association is typical of IBM's cloud customers. Last fall, USGA turned to IBM for business continuity services, including daily backups of its business-critical servers and, in case of emergency, e-mail recovery. Both services are hosted in IBM data centers. Jessica Carroll, USGA's director of IT, says the business continuity services have improved USGA's readiness if its own systems should fail during one of USGA's 13 annual championship golf tournaments.
With years of experience, hundreds of customers, partnerships with Google and Amazon, and growing portfolio of on-demand services, why doesn't IBM get more respect as a leader in the cloud computing market? Because IBM isn't a player in the hottest areas of the market: platforms as a service, infrastructure as a service, and Amazon S3-style storage as a service. Customers looking to deploy virtualized IBM software stacks (DB2, Informix, WebSphere Portal) in the cloud, must turn to Amazon, which offers it, not IBM, which doesn't.
IBM hasn't revealed plans to offer those types of flexible, pay-as-you-go cloud services, while Sun, with its Sun Cloud and Open Cloud Platform, is moving in that direction. That explains why some people see an IBM-Sun combo as being a good fit. At this point, however, that's speculation, not a roadmap.