HP is developing hardware, software and electronic service technologies that will enable users to build powerful, highly scalable and flexible data centre architectures at a fraction of the current cost.
Under the company's Always-on Internet Infrastructure strategy, users will increasingly be able to share computing resources across multiple applications, rather than having to dedicate specific resources to specific applications, as is the case with today's data centres.
Under the new initiative, a server that runs one application in the morning can be freed up to run another one at night, with all of the reprovisioning work being managed from a central console.
Applications will be able to use the resources of multiple servers within one data centre or across multiple data centres. HP said it would also provide usage monitoring and management technologies, which would allow companies to adopt a pay-for-use model.
Andy Ledbetter, a HP spokesman, said technologies such as the HP Hyperplex server consolidation platform and the "instant capacity on demand" feature on high-end Superdome Unix servers were examples of enabling technologies that the company would deliver in the next few months.
"As a user, this is generally a direction I'd like to see HP move in," said Edward Witkow, the director of IT at Metaldyne. "The utility-like computing, which HP is talking about, will lower costs and improve the predictability of IT budgets."
Striking a note of caution, however, Witkow warned there was much work to be done if HP's plans were to fully benefit the industry. "There are issues that HP needs to address first," he said. "For example, HP's instant capacity on demand technology allows users to quickly add capacity when needed. So far, however, there is no way for users to decrease that capacity when it is no longer needed."