24 January 2006 The decision to move to a utility-based model of computing is an economic inevitability because current methods of delivering IT are too inefficient and provide too little competitive edge, delegates at Information Age‘s Effective IT Summit were told today.

Launching the Effective IT Summit, Harvard Business Review contributor and author Nicholas Carr used his keynote address to warn delegates to start planning for a transition to utility computing that will take a decade or more to complete.

IT is fundamentally a resource for doing business, and therefore its development will be governed by economic rational, Carr explained. “The supply of any business resource will gravitate towards the most economically efficient model. Any company wants to operate as cheaply as possible.”

Certain types of resource – such as electricity – lend themselves to being delivered as a utility service, said Carr. So while factories originally had their own power generators, once an economically viable utility model was developed, businesses shifted over. The same evolution is underway in IT.

“When you consider whether IT could be delivered as a utility, you have two questions to consider: Is the current model of supplying the resource the most efficient we can imagine, and is it technologically possible to deliver the resource as a utility?” he said.

Current utilisation rates of servers, PCs and storage systems are incredibly low, Carr highlighted. Moreover, the vast majority of an organisation’s IT spend goes on ‘keeping the lights on’. Vast amounts of money are being poured in IT infrastructure that adds no competitive advantage, he said, so, it is inevitable that companies will emerge that deliver that infrastructure as a utility service.

The question then becomes whether it is technologically possible to develop a utility model, where economies of scale and other efficiencies mean that commodity resources are delivered to business users at the lowest possible cost.

Virtualisation, grid computing and web services all pave the way towards utility computing, said Carr. “Critical is automating IT. It’s a great irony that in the last 20 years, IT has automated almost everything other than itself.”

“Of course, utility, grid computing, software as a service and web services have all been adopted as marketing jargon by the IT vendors: that rightly brings out some scepticism,” said Carr.

“The transitional from the current client-server model to utility computing may take some time; data centres are not going to be mothballed overnight. But as these technologies advance, there are significant opportunities to drive down the IT costs in business.”

“It is only those companies that achieve the transition to this vision first that will be successful from an IT perspective,” he said.

Not everyone agreed with Carr. For one, Doug Richard, chairman of technology angel investor Library House, argued Carr’s analysis was fundamentally ‘flawed’.

“[Carr] seems to talk about the end of innovation, but there will always be innovation. Information technology is not electricity where the unit is an electron; in IT every byte is different. It is what is transported that matters,” said Richard.