Profile: Roger Needham - Bill's boffin bares (nearly) all

Microsoft's British research chief was happy to show Dana Rubin around the new lab ... as long as she didn't ask about the share options

THERE ARE no oscilloscopes with peaking and troughing green lines, no screeds of paper covered with zeros and ones, no humming mainframes. In fact, at Microsoft Corporation's state-of-the-art new research laboratory in Cambridge, there is no equipment at all, except for some ordinary-looking desktop computers.

"Eventually, we will have a room for doing experiments with physical things," explains Roger Needham, the 63-year-old expert in computer security who was hired by Microsoft last summer to head the new lab. What sort of experiments? "Well, we don't really know yet," Needham adds.

It has been nine months since the US software giant announced plans to spend $80m (pounds 48.9m) over the next five years setting up a centre for computer science research in Britain. During that time, Needham has already lured some of Europe's best research talent to work for him, but so far he has only vague notions about what they will be doing.

"The way you get very good ideas is you get the best people and give them the freedom to think and experiment and let their imaginations rip," says Needham, who has spent a long and prestigious career in computing at Cambridge University.

Microsoft's lab at Cambridge is the company's first overseas offshoot of its seven-year-old research centre at Redmond, Washington, and further proof of the company's intention to create the world's largest computer science research operation.

At the centre in Redmond, Microsoft Research has a dozen research groups studying areas ranging from speech technology to cryptography to decision theory. In the next three years, it intends to increase its overall number of researchers from 245 to about 600.

Microsoft's plan has elicited a chorus of praise around the world at a time when large corporations are less willing to take risks and invest in basic scientific research - that is, research that is not product-oriented. Companies such as Xerox, IBM and AT&T still invest heavily in basic research, but corporate commitment in general is nowhere near as high as it was from the 1950s to the 1970s.

Even Microsoft's top executives acknowledge that they are expecting commercial returns from the research they are supporting. "Yes, it will benefit us. This is not an altruistic activity," pronounces Rick Rashid, vice-president of Microsoft Research."We expect to see results in the sense that Microsoft can bring new and exciting technologies to market."

To continue growing and to establish control over software infrastructures on a global scale, Microsoft must become a technology leader.

By studying an arcane subject such as statistical physics - how complex systems operate in the natural world - Microsoft researchers hope to gain a better understanding of how large networks and clusters of computer systems work. Such knowledge is absolutely critical for a company that is increasingly selling its wares to other companies rather than to individuals.

Or, for example, through the study of natural language processing, researchers are trying get computers to understand, analyse and generate language like human beings do. At the Redmond research lab, studies in natural language have already yielded practical results in the form of a new grammar- analysing function incorporated into the company's Office 97 software.

The company's ultimate goal is to develop new and better ways for people to interact with computers, home appliances and entertainment.

"They have this asset that no one else has - 200-plus million users - and the ability to influence their spending by selling new generations of their product," says Walter Winnitzki, an analyst with PaineWebber in New York. "Once you have a product breakthrough that 200 million users can use, the research pays for itself over and over again.

In truth, the $80m that Microsoft is spending in Cambridge is not a lot for a company that, this fiscal year, is spending $2.6bn on overall research and development.

"It's chump change," says Jonathan Steel, a business technology consultant at Bathwick Group in Bath. "The issue is more what they are going to end up with, which is 40 top-grade academic minds doing work on Microsoft note paper. Even if they don't get solid products out of it, it is still good public relations."

Microsoft's executives are eager to convey the message that the company is making a commitment to fundamental research at a time when federal financing for basic research in the US is at a standstill. Aside from the $80m, Microsoft has also given $16m to a venture fund for start-up technology firms in the Cambridge area.

This, too, could have a commercial pay-off for the company. "If Gates can put some money into a VC fund and hit one or two interesting companies, then clearly it would be of great value to him," says Steel.

The William H Gates Foundation, funded by Microsoft founder and chairman Bill Gates, will also donate $20m towards a new building a mile-and-a- half west of central Cambridge that will serve as the lab's permanent home. But why set up a lab 5,000 miles from Redmond in the first place?

Microsoft officials say they needed to branch out internationally to attract researchers who wouldn't want to relocate to the west coast of America. As the only overseas location where Microsoft could tap into a world-class science community, with the bonus of an English-speaking environment, Cambridge was a logical choice.

Cambridge University also happened to have sentimental significance to Nathan Myhrvold, Microsoft's chief technology officer, celebrated polymath, and the driving force behind Microsoft Research. As a post-doctoral student, Myhrvold left a fellowship under Cambridge cosmologist Stephen Hawking to become a software entrepreneur.

In December 1996, at the Westin Hotel at San Francisco airport, Needham met with Myhrvold, Rashid, and computer designer Butler Lampson, as they outlined their plans for the British lab.

"So what do you want me to do?" Needham recalls asking them.

"Go out and line up the best people," they said.

Needham said yes. Among the first to be hired was Derek McAuley, a specialist in networking, distributed systems and operating systems and a computer science professor at the University of Glasgow.

Then there was Chuck Thacker, an American who was a co-inventor of Ethernet and chief designer of Xerox's Palo Alto Research Center computer that inspired the Macintosh, which in turn inspired Microsoft Windows. Thacker was hired from Digital Equipment Corporation's systems research centre in California.

The team has been joined by Luca Cardelli, an Italian specialist in programming language theory, Dieter Gollmann, an expert in security from Austria, Ralph Sommerer, from Switzerland, who has a PhD in computer systems from Zurich, and Stephen Robertson, a British expert in the field of information retrieval.

In the coming 18 months, Needham expects to have a grand total of 40 researchers on board. He insists it is not money that is drawing researchers to Microsoft, because the salaries that he is paying are not "terribly high". Senior researchers at Microsoft Research in the US are reported to earn close to what they would make in academia.

But what about those golden stock options that are part of the company's legendary compensation package? It is because of those share options that Microsoft has created more millionaires than any other company in the world.

Needham hesitates. "We have a few no-go areas for talking to people like you," he says, "and this is one of them."

Copyright: IOS & Bloomberg