Just when you thought it was safe to go back into the water, the debate over the definition of cloud computing has reared its ugly head again.
The topic apparently kicked off a panel discussion at the recent VoiceCon event, with seven different answers from seven different panelists (ranging from “anything related to a computer in the public network,” to a “network-based communication utility,” to “an evolution starting with Centrex, IP-Centrex”). And while the AT&T spokesperson on that panel said the company wasn’t defining “cloud,” Joe Weinman, VP of Strategy and Business Development for AT&T Business Solutions, recently took a stab at doing so in a post at GigaOM, defining CLOUD as a “Common Location-independent Online Utility service, available on-Demand.”
The most interesting discussion of what cloud computing means came from the VC community, though. According to James Urquhart in his “The Wisdom of Clouds” blog at CNET, at least one VC firm, Sierra Ventures, is giddy over “cloud” and wants to see a lot more of it — despite the fact that its overuse seems to have rendered it permanently nebulous. However, Urquhart, who prefaced his presentation of Sierra Venture’s cloud stance with a suggestion that the term is overused, was called out himself for stretching the definition to fit a particular purpose.
Urquhart, whose day job is market manager of Cisco’s Data Center 3.0 initiative, has been doing analyst briefings for Cisco’s InterCloud strategy (learn more about it here), and the strategy’s terminological tangle (public clouds vs. private clouds vs. the InterCloud) left Gartner’s David Smith questioning if we’ll ever get past confusion over what is meant by “the cloud.” Urquhart attempted clarify Cisco’s vision via a post on the company’s Data Center blog.
All of this brings us to new research from McKinsey & Company. The research, which asserts that most big companies would lose money in a move to cloud computing, omitted a definition for that term. The meat of the research is this: McKinsey compared using Amazon EC2 to running a traditional data center, and it found the cost of running data-center functions would more than double. But in such a calculation, the report ignores the existence of private (or internal) clouds.
Through a mix of federation, virtualization and dynamic management software, many companies are building private clouds — essentially shared, service-oriented infrastructures — and saving money doing so. By implication, then, McKinsey’s definition of cloud computing does not include private clouds. So, when it says that cloud computing will lose big companies money, it means public cloud offerings will lose big companies money.
And I thought we had left this debate behind (sigh).