Google, MS, Sun team on Net services lab

Google Inc., Microsoft Corp. and Sun Microsystems Inc. have set aside their differences long enough to jointly help create an Internet services lab along with researchers at the University of California, Berkeley.

The Reliable, Adaptive and Distributed systems lab, or RAD Lab, will focus on developing technology for small teams or individuals to create Internet services, according to a press release issued Thursday.

The three companies will provide US$7.5 million in funding over five years. Industry, including the three founding members as well as other companies making smaller contributions, will provide 80 per cent of the funding, the statement said.

Other funding will come from the U.S. National Science Foundation and from the University of California’s Discover and Microelectronics Innovation and Computer Research Opportunities grant programs.

The RAD Lab will explore alternatives to the traditional software development model, in which work is completed in orderly stages, with infrequent, well-tested upgrades.

Code for Internet services is now being modified on the fly as the services are scaled up to accommodate millions of users, but this type of development requires a large technical support staff. The RAD Lab will focus on eliminating the need for large teams to develop such services. One technology that can help make this possible is statistical machine learning, according to a statement attributed to David Patterson, UC Berkeley professor of electrical engineering and computer sciences and founding director of the lab.

Any software emerging from the RAD Lab will be made freely and openly available, with the source code distributed using the Berkeley Software Distribution license, the statement said.

Leading the RAD Lab will be five UC Berkeley professors in electrical engineering and computer sciences, as well as an assistant professor of computer science at Stanford University who is expected to join UC Berkeley in July 2006. They will start out with 10 computer science graduate students; that number is expected eventually to grow to 30, the release said.