IBM wants some of that Web 2.0 mojo. That is what is behind its announcement today of Blue Cloud, a set of “cloud computing” offerings that will be available to its corporate customers in the first quarter of 2008. Of course, cloud computing is just Web computing by another name. It implies massive server farms, massive storage, and the ability to support Internet-scale applications and usage patterns.
Amazon, Google, Yahoo, and Salesforce.com.com are all examples of cloud computing already available to consumers and businesses today. These are mostly in the form of specific applications, but Amazon offers its suite of Web services, which are a collection of generic cloud computing offerings—computing cycles, storage, communications. Even Salesforce.com offers its cloud computing infrastructure to other companies through its AppExchange. IBM should not have any trouble competing.
Blue Cloud is being billed as more of a distributed computing architecture than what you find in most corporate data centers. It is based on an open-source project called Hadoop that manages computing resources across large clusters of computers. Hadoop includes an open-source version of MapReduce, the same software Google uses to efficiently distribute its computing chores across its servers around the world.
So IBM is basically bringing this massive-scale computing architecture to its corporate customers. That will be good for corporate applications because this sort of distributed architecture lends itself to Web 2.0 apps, which are already invading the enterprise. The question remains as to how exactly IBM is going to implement Blue Cloud. Will it offer it on its own hosted server farms around the world or teach big company CIOs how to build their own mini-Googles across their own data centers?
My guess is it will probably be a little bit of both.