Who's Doing What

IBM, HP and Sun may be the early frontrunners in utility computing, but where are Microsoft and Dell? We all know they can't be far behind. A first hint that Redmond has plans around the "virtual data centre" came recently, in comments as it gears up for the release of Windows Server 2003. And, of course, the hype surrounding its .Net initiative - largely focused in Web services - is almost deafening. Let's hope it gets understandable; ask 10 people what it is, get 10 answers. (One can only hope Big Blue continues to raise the industry bar in educating business about IT.)

And Dell? It can't not be a player, but its model doesn't call for getting in early on new trends. It cleans up later, once the volume play kicks in. (Case in point: the stellar quarterly results it just reported were pumped up nicely by new storage offerings, of all things, a game it only recently entered.) Sun's N1 initiative is about making servers, storage and network equipment work better together, and is initially focused on "virtualising the data centre". That means pooling disparate servers and storage systems and making them seem a single, large super-server. As such, N1 is competing with management approaches from IBM, variously called eLiza and Autonomic Computing, as noted above, and with HP's Utility Data Centre and Adaptive Infrastructure initiatives.

HP claims to have had the vision for utility computing some 20 years ago. Last US autumn, it launched its Utility Data Centre. (No explanation why it took so long.) HP says companies can adopt utility computing in one of two ways: outsource the entire IT operation under a contract that charges the customer only for the server capacity, storage, bandwidth or applications it uses; or get on-demand access to storage and server capacity that scales larger as the customer's needs increase. How does that differ from IBM's pitch? Not much would be a good answer.

At IBM, initiatives have been in the R&D phase for many years that it claims laid the groundwork for its On-Demand Computing push. For example, the company has built computing grids, pooling the collective data processing and storage capabilities of large networks, going back decades. And, for more than a year, Big Blue has also been pushing its Autonomic Computing initiative, aiming to build those machines that can diagnose and repair their own problems.