When the call came from Hurd, Mott was faced with his greatest challenge: to take a disparate IT infrastructure and simplify it without reducing its scope or interrupting business in the interim.

"One of the challenges is that, even though most of us look at the possibilities and capabilities of technology as it is today, the underlying challenge is that most CIOs and organizations don't have the benefit of today's technology in most of their operation," he points out.

"If you look at most IT organizations, there are some old technologies running many applications, and those technologies have to be supported. There are cost implications of that as well as limitations. Obsolete technology is probably the biggest barrier to them being as effective in helping the business as they should be."

Mott is no supporter of the adage "if it ain't bust, don't fix it" as an argument for hanging on to legacy systems. "The other part of the argument is that old systems do break and you have to fix them -- that can be costly," he says.

"Also, they don't grow. It has a 'layering' effect because if you don't get rid of things and you just keep adding layers, it becomes a big part of the cost structure. If you look at the cost of IT over the past 15 years, it is growing faster than the economies of the world. That's not a good thing. All of the people and support costs go into keeping these systems going. Because we don't get rid of things, it is keeping us from being as cost-effective as we could be."

At HP, this issue has instigated a major project to reduce costs on a more-for-less basis. Mott's team is overhauling the infrastructure and reducing its number of datacenters from 85 to three paired installations. The server count is being cut from 21,000 to 14,000 by using virtualized consolidated BladeSystem servers.

"The path we set out on just over two years ago is what we term 'IT transformation'," he says. "We found ourselves having about 5,800 applications. We assessed it and decided we should only have about 1,400 applications. We have a lot of duplication in all ages of systems. We've got a need to simplify from the standpoint of getting rid of a number of applications that may do the very same function for different parts of our business, and getting the whole area on the same application.

"There's a lot of business change associated with that and we're having a major refresh of technology. We were using older technologies that were becoming brittle because of their age. There are a lot of mechanicals on these systems and mechanicals break down. So we need to refresh the technology base and do it on the latest servers and storage.

"We can rework our network and, at the same time, we have a need for better information. We have lots of data in the company and we have lots of data marts -- more than 750. The transformation gives us an opportunity to get better, consistent information about the company and be able to look at that across our different businesses in a timely fashion."

The database change brings Mott back to his old area of expertise.

"We have a strategy around enterprise data warehouse, which is being implemented on our own Neoview technology. We started implementing it in April 2006, working closely with our product team. We're really driving to get to a single data warehouse, which no one does today. There are large data warehouses, but no one I know of actually has an enterprise system holding all the company data."