Hewlett-Packard this week demonstrated new software that potentially could let corporate users take advantage of grid computing to run business applications. But a lot more pieces need to fall into place before that becomes a reality, HP and analysts said.
Grid computing technology is designed to let users harness the computing power of numerous systems. For example, grids made up of thousands of linked computers are being used to run large engineering and scientific applications, mainly by universities.
HP's demonstration at the GlobusWorld grid computing conference in San Diego is a step toward delivering the same capabilities in a commercial application environment, said Sharad Singhal, a principal engineer at the company's HP Labs research unit.
HP showed how users could link its Utility Data Center software to a grid computing environment, enabling data center resources to be used by grid users and vice versa. Singhal said HP hopes to eventually be able to support gridlike networks on which business users looking to increase the computing power available to them will be able to tap into a vast pool of shared resources.
As part of the demonstration, HP also showed a tool that uses drag-and-drop techniques to let users define and advertise their resource requirements in such a network.
But increased security requirements and the complexity of adapting commercial applications to take advantage of the grid computing model means it will be quite some time before the approach is used in business environments, Singhal said. "These things will begin to take shape in the next year to a year and a half," he said. "But how quickly they become a commercial reality is anybody's guess."
Jonathan Eunice, an analyst at Illuminata, said the much higher I/O and application performance requirements of commercial data processing environments will also limit grid computing's appeal to business users.
"We need to stop talking as though the traditional database application is going to be distributed over a grid," Eunice said. Instead, he said, he expects to see companies adopt the technology to power large analysis applications involving corporate data.