Now's the time for utility computing

If you ask an IT manager about the prospects for utility computing, odds are the manager will respond with something like, "Sounds nice, but I have real problems to deal with here." But like objects in the rear-view mirror, utility-based computing is closer than it appears.

For it to have any commercial potential at all, utility-based computing had better have some near-term applicability and some potential for significant benefits--enthusiasm for fuzzy, utopian visions of computing from actual customers is certainly in short supply these days. Proponents of utility-based computing have had a hard time avoiding the tendency to speak in "gee whiz" terms. Although well-entrenched in scientific and technical circles, elements of utility-based computing, including grid computing, remain a puzzle to many commercial enterprise IT managers. IBM has only recently begun to back up its expansive e-business on demand vision with tangible services such as the recently introduced LinuxOnDemand service offering. This same challenge also exists in the realm of Web Services. In acknowledging the difficulty of spurring near-term investment for its .Net strategy, Microsoft has recently begun emphasizing the near-term possibility for cost savings as opposed to the long-term potential.

While grid computing has become synonymous with utility computing, it is but a part of a much broader canvas. As explored in Hurwitz Group's recently published Market Map, "Utility-Based Computing," utility-based computing is a metaphor that encompasses a variety of technologies including server virtualization, automated provisioning software, and blade servers, which together comprise the next-generation data center. Hurwitz Group's definition of utility-based computing includes both the technologies that enable utility-based computing as well as the modes with which they can be delivered, including software-as-a-service and variable-cost, variable consumption outsourcing models. Whereas much of the current advertising has focused on utility services, utility computing, like Web Services, is both internal and external in its orientation. As is true in the case of Web Services, it is the internal orientation that is gaining the most traction in the short term. A principal focus of utility-based computing is the emergence of tools and technologies that can facilitate server consolidation and higher utilization of server infrastructure so that commercial IT organizations can achieve greater value out of existing investments as well as future investments. In this context, utility-based computing is ultimately about making better use of all computing resources.

To be sure, significant gaps remain in the next-generation data center scenario, including an effective end-to-end management paradigm. But the economics of managing current server infrastructure are so punishing that data center managers have little choice but to at least investigate them. Moreover, if activity in the venture capital community is to be trusted, utility computing continues to be a major focus of investment. For example, blade server pioneer Egenera recently gained $44 million in additional funding while recent data center virtualization start-up Topspin received a $30 million haul. Meanwhile, additional automated infrastructure provisioning players, including the recently announced ProvisionSoft, continue to emerge. More importantly, the large industry players including IBM, HP, Microsoft, and Sun are making utility-based computing the focal point of their investment schemes in full recognition that utility-based computing is emerging as the new battleground in which future dominance of the computer industry will be established. Significant announcements, partnerships, and even acquisitions can be safely anticipated. More important still, however, is how the nascent utility-based computing movement can provide near-term relief to over-burdened IT managers. This potential will reward near-term evaluation, and ultimately, customer investment.

"The Time for Utility-Based Computing Is Now"
Originally published July 26, 2002 by Hurwitz
By Bill Martorelli

How is your organization implementing utility-based computing? TalkBack or send e-mail to us.