HP's Focus on Utility Computing

The pioneering computer firm tries to anticipate enterprise demand in 2010.
Posted October 27, 2005

David Needle

HP Labs isn't the only R&D facility in the Silicon Valley. In fact, it isn't even the biggest one. But it counts a string of innovative breakthroughs to its name.

And despite recent management changes, and the distraction of numerous acquisitions, the prototypical Silicon Valley firm, which started in a garage, has kept its investment in R&D high with some $3.5 billion slated for this year, of which about 5 percent is used by its labs.

HP Labs in Palo Alto employs 400 people and is one of six labs worldwide with a total of 700 employees.

''We're very healthy and we innovate to solve real customer problems,'' Rich Friedrich, director of HP's Internet systems and storage laboratory, told internetnews.com.

''Some other labs have gone through a kind of pendulum of speculative research and then switched to a very short-term payback focus, but we've always been very focused on innovating to solve real customer problems. We're not ivory tower, we meet with customers all the time.''

A project launched at the labs three years ago, Enterprise 2010, tries to forecast what the demands of tomorrow's data-center customer will be and how best to address them.

One big aspect of that is HP's work in utility computing -- the idea of pay-as-you-go access to computing resources.

Other companies, notably Sun Microsystems, with its grid utility model, are pushing a similar scheme. Sun is essentially saying ''come and get it'' with dollar-per-hour-per-CPU pricing, while HP is working with a few select clients to develop the program.

''We're very early on with this idea of offering temporary or peak capacity on a per-use or as-needed basis,'' Frank Gillet, principal analyst at Forrester Research, told internetnews.com. ''There are a handful of early experimenters, but the software providers aren't necessarily ready to do it yet; a lot of work still remains.''

This article was first published on GridComputingPlanet.com. To read the full article, click here.

Comment and Contribute


(Maximum characters: 1200). You have characters left.