Sun Shifts R&D Focus To Horizontal Computing

Sun Microsystems' R&D efforts over the next several years will emphasize emerging technologies for service architectures, Sun CTO Greg Papadopoulos said during a Sun Labs’ open house event Friday.

Though Sun will still dedicate resources to maintaining products for traditional corporate computing, the thrust of the company's research efforts will be focused on developing scalable hardware and software for companies with business models akin to Salesforce.com, Amazon.com and Google.

"It's an amazingly difficult space to navigate," Papadopoulos said.

Papadopoulos, who spoke to the press and analysts Friday at a Sun Labs' open house, said Sun's R&D will "likely change" as a result of the company's restructuring plans, announced earlier this week, but declined to detail specifics. Sun’s annual R&D budget has historically been about $2 billion per year.

Sun, which has struggled over the past several years to remain profitable, said Wednesday it would cut as many as 5,000 jobs, or 13 percent of its workforce, over the next six months.Sun Labs, the division that investigates advanced technologies, accounts for about 2 percent of the current R&D budget and that will not change as a result of the restructuring, said Papadopoulos. The mix of projects, however, may change, he added.

On the corporate IT side, Papadopoulos said applications such as server consolidation, virtualization and information life-cycle management will continue to drive sales. But, he said, more interesting advances will be had in how to help large service providers manage networks of systems laced together in clusters or grids and how to develop operating systems standards for horizontal computing.

"This is an interesting time to do that kind of R&D alignment in the company," said Papadopoulos. "We haven't really done that in the past five years."

Sun Labs showed off several projects intended to boost the performance of clustered computers and add value to service provider networks. Papadopoulos talked about, but Sun did not demonstrate, Project Maxwell, designed to "challenge the assumptions” of how computer memory systems work. The project is looking at building virtual memory as objects, he said.

A project to optimize database architectures on clusters and another to use tiny wireless sensors that can be easily programmed in Java and talk to service provider networks were among the technologies demonstrated at Sun Labs' open house.0