IBM Unveils Blue Cloud - What Data Would You Like to Crunch?
IBM announced this morning that it will be rolling out cloud computing services for corporations in the Spring of 2008. More than just a beautiful turn of phrase, cloud computing is a paradigm that leverages a distributed architecture to carry out massive processing tasks online, instead of on a single computer. The program, called Blue Cloud, is set to compete at least indirectly with Amazon's Elastic Compute Cloud, or EC2.
In theory, cloud computing will usher in an era wherein innovation based on massive data processing becomes affordable for almost any company, no matter how small. The possibilities are exciting; I think of the huge benefits we've seen from the network effect of users aggregating their activities online and the thought of leveraging a related paradigm for processing power seems fantastic. I hope we'll get to see the direct application of such an approach, but cloud computing could remain an activity of big, stuffy organizations only. Any semantic web companies interested in both heavy processing and avoiding investment from In-Q-Tel, the CIA's investment arm that's ubiquitous in the semantic web sector, might appreciate improved cloud computing options. IBM's hardly got a perfect record historically, either, but nobody's perfect.
IBM announced last month a collaboration with Google to promote cloud computing in academic institutions. Though from the consumer perspective working with Google sounds promising, Google has seen very limited success with large organizations. Similarly, RedMonk analyst James Governor argues that the real story behind Blue Cloud is that "IBM is miles behind the real players in this space and needed to win some breathing room." "I see no evidence IBM gets Cloud Computing, " Monk said. "IBM is addicted to its own scale, so it doesn't fully understand web effects. What's Blue Coud? Enterprise EC2/S3 but without customers and developers at this point."
Amazon's showcase customers for EC2 are mostly tiny, random startups, so it's hard to know the extent of adoption, but the potential is big. The market for time on single Super Computers is substantial, so in theory cloud computing should see a lot of demand as well.
Is cloud computing set to become commoditized? That probably goes without saying. What direction will it move in then? I can't help but think that IBM is uniquely prepared to add a layer of business services on top of their cloud. It's mornings like this that I wish I was more engaged in the enterprise space - this sounds like a whole lot of fun.
Larry Dignan at ZDNet, a writer far more informed on enterprise matters than I, says that "It’s still early to ponder the enterprise cloud possibilities, but it can’t hurt to mull over the issues. The cloud will be hovering over your corporation shortly." There's also some informed discussion over at DataCenterKnowledge.
Interested in other large scale, futuristic work at IBM? Check out this post here on Read/WriteWeb earlier this year from guest authors Martin Wattenberg and Fernanda B. Viégas, two researchers behind IBM's Many Eyes program.
Comments
Subscribe to comments for this post OR Subscribe to comments for all ReadWriteWeb posts
I'd like to see more wide spread adoption of some of the cloud computing projects than anyone can join like SETI @ home (looking for ET in radio noise from space) or folding @ home (cancer research.) I think these programs should come installed and enabled as the default on all new computers.
Then, you could turn them off if you wanted, but otherwise, all that work would get done for the greater good.
While IBM's announcement may focus on parallel processing and Hadoop, cloud computing goes well beyond number crunching.
Most users of Amazon's EC2 or 3tera's AppLogic use these cloud computing services as replacements for traditional data centers. Simply supply a credit card and you have a full blown data center at your disposal ready to map the human genome - or to run your web 2.0 app.
If you've ever had an older computer scientist for a teacher, friend or company, they would have told you how they had to "book" time with the "mainframe" in the "computer room" (my fingers are aching now). Usually to crunch some data or prove some theorem.
Now they can buy time in order to achieve same. The commodity of physical space and computing power has changed to that of time, money and computing power.
mad props to marshall for remembering and linking to ibm and the holocaust. it is important to never forget.
you really need to be able to trust your cloud computing service, which is why i think a lot of these cloud companies need to be way more customer service oriented. amazon's service is great and very innovative, but i dont think any player in this space is customer service oriented enough.