)? "It's all about possibilities," he says. "The reason I went into computer science is that I wanted to create something that hadn't been done before."
As professor in this field at Carnegie Mellon University in the 1980s, Rashid was lead developer of a computer operating system that moved data more efficiently than other programs. The OS, called Mach, was freely released to developers to use as they wish, and the core of the technology is still active inside Apple Computer Inc. (AAPL
) 's OS X software for the Mac, among other places.
Microsoft hired Rashid, 54, from Carnegie Mellon in 1991, and he now runs one of the most ambitious basic-research units in the business. As senior vice president of research, Rashid manages more than 600 research scientists scattered among labs in Bangalore; Beijing; Cambridge, England; San Francisco; Silicon Valley; and Redmond, Wash. Despite the sprawl, Rashid tries to model this operation on a university lab -- though most academics, however, would probably envy the $250 million a year that Rashid has to throw at projects on the frontiers of information technology.
Rashid's work makes up a tiny piece of Microsoft's overall $6.2 billion research and development operation, but it's one of the most important parts. Mere incremental improvements to Microsoft products aren't the charter of Rashid's group. Instead, he pushes his staff to go after the big challenges: imbuing machines with the ability to understand language exactly as it is spoken, ironing the inefficiencies out of computer networks, generating sharp computer graphics a viewer might mistake for reality -- and yes, tackling the next, industry-altering leap in operating systems. "There isn't any constraint," Rashid says. "Research has no schedule."
All of this requires changing the traditional metrics behind corporate research. Rather than measuring researchers' performance based on the amount of work that gets transferred into the company's products, staff members are judged on the number of important papers they unveil in technical journals or at conferences. His unit is a magnet for talented PhD scientists. Several Microsoft researchers, including Rashid, are members of the National Academy of Engineering, something of a geek hall of fame. Many of the team are "people who want to change the world," he says. "They have an evangelical zeal."
As with all basic research, projects carry a high risk of failure, so staff members don't assume they'll see their hard work built into products. Still, the chance to put an invention in the hands of Microsoft's hundreds of millions of users is a heady incentive, and it's striking how often Rashid's team pulls it off.
Responding to critics who carp that Microsoft just copies others' best work, Rashid can reel off dozens of strides his team has made -- like the grammar checker in Microsoft Office, and ClearType, a display technology that produces crisper text resolution in Windows XP. Researchers also came up with spam filters that block unwanted e-mails in MSN Hotmail, a free feature that stops 3.2 billion messages a day. The steady anti-Microsoft harping is "one of those things where people don't think it through," Rashid says.
Rashid's own innovations didn't end with Mach. Among other things, he has developed video games. His first title was Alto Trek, a sci-fi romp released three decades ago. In 2000, Microsoft's games division came out with a sequel, Allegiance. Games are a definitely a gas, Rashid says. But it's the science that lets him boldly go where no one has gone before. By Jay Greene