View from the Fens Windows

Roger Needham is the man in charge of Microsoft's Cambridge is the man in charge of Microsoft's Cambridge millions. Tony Durham asks how he will spend them

Microsoft is not the first computing company to set up a lab in Cambridge. In the past 20 years the city has attracted Xerox, Schlumberger, Logica, Olivetti and Oracle, not to mention all the local Cambridge Phenomenon outfits in timbered riverside lofts or steel-framed science park units. By 1996, Cambridgeshire had 892 high-technology companies employing 28,000 people.

These firms did not come for the architecture or the punting. They came for the brains. Their researchers did joint projects with the university computing laboratory or other local centres such as the Medical Research Council's Applied Psychology Unit. Some prominent researchers such as Andy Hopper have shared their time between university and industrial labs. But Microsoft's arrival seemed to mark a new stage in Cambridge's relationship with the computing industry.

Microsoft is huge, aggressive, brash, populist and self-celebrating. When it decided to set up a Cambridge laboratory a culture clash seemed likely. Microsoft and its chairman Bill Gates are easy targets for snobs. Gates may own the digital rights to much of western culture, and he can change the old masters on the wall of his Seattle mansion by pressing a few buttons. But that is not a video screen on the wall of King's College chapel, it is a real Rubens. Welcome to Cambridge.

The news that Microsoft was setting up a Cambridge lab broke in May 1997, a few days before we learned who would be running it. Roger Needham, professor of computer systems and pro-vice-chancellor of the university, was an obvious candidate but definitely not the kind of person who would quietly take orders to develop Windows 2000 or whatever product was proving a bit too tough for the teams of hard-working young developers on Microsoft's campus in Redmond, Washington.

Microsoft's decision to invest Pounds 50 million in the Cambridge lab over five years was a small victory against the brain drain that has drawn 10,000 Cambridge University alumni to the United States. Though the city now hosts Europe's greatest concentration of high-tech activity, since 1950 there have been calls to limit its growth and protect its character and quality of life. Fears that Cambridge will lose its sparkle and become an outpost of California's Silicon Valley are voiced in a report published this week by the consultancy Analysys. One of the report's seven authors is Sir Alec Broers, vice-chancellor of the university. Another is local entrepreneur Hermann Hauser, head of the Amadeus Capital Fund that has invested in five United Kingdom high-tech companies so far. Last year Microsoft expressed its commitment to support Cambridge-area companies by putting Pounds 5 million into Amadeus.

The report, Cambridge 2020, is funded by the telecommunications company Alcatel, and argues that Cambridge can overcome the physical limits to growth, such as land shortage and traffic congestion, by investing in communication networks. Cambridge could become a "wired city" on the model of Singapore or Malaysia's Multimedia Super Corridor.

Microsoft was founded in 1975 and after 16 years of phenomenal growth, decided it must begin doing research. "The short way of saying why is that it sets the agenda for the company," says Needham. "If you are a small company can pick your technology product agenda out of the atmosphere around you. If you are a big company you cannot. You have to make it for yourself and that is what research does.

"So they started off research at Redmond with a view to setting an agenda. A few years later, they were very happy with what had happened. They had got up to about 130 people and every Microsoft product was influenced for good by research. Now this was not because the research lab had been told to go and develop products. They had been told to go and do interesting things and keep in touch. And they kept in touch to the effect that there have been very good results for the products."

Microsoft decided to triple its research effort and hire another 300 people. But not every computing researcher in the world wanted to work in Redmond.

"After some chain of argument that I was of course not privy to, they decided to have a lab in Cambridge, and asked me to run it for them. I think they knew very well that I would probably say yes," explains Needham.

Before he took the job he asked what he was meant to do. "Go and hire the best people there are and have them do what they are good at," was the reply. Nathan Myhrvold, now the company's chief technology officer, told him: "If every project you start succeeds, you have failed."

It was not the Windows 2000 job. It was real research. "Add to that the merits of being 62, which means that the concept of career risk is nonexistent, and it did not take me very long to come to the decision in principle - about a quarter of a second."

Microsoft Cambridge Research now has city-centre premises and about 15 staff. Needham plans for lab staff numbers to grow to about 40 people. There is a nascent plan to move to a shared site with the university computing lab in three or four years.

"The areas we are working in are determined by who we can get," Needham says. "Rule one for running a good research lab is, you have to have high-class people. It is no good saying 'we must work on area X and OK, this guy is plausible, let's do it'. That way lies ruin." His strategy is to get mid-career European researchers with international reputations, who will attract others.

Programming language theory was not on Needham's research agenda until the Italian expert Luca Cardelli joined the team, attracting two other specialists in the subject. Neural networks looked interesting, but Redmond already had a group in the area. Needham asked whether his lab should get involved in the subject. "They said if you can get X or Y do it, otherwise leave it to us. So we got X. We have not got Y yet." X was Chris Bishop, from Aston University. The unnamed Y is being watched for signs of job dissatisfaction.

If he has time, Needham will do some work in his own subject, security. "Running the business and being pro vice-chancellor of Cambridge I have usually been in both my offices by eight in the morning. It does not leave me a great deal of time for pursuing my own interests," he explains.

His chief lieutenants are Derek (Mac) McAuley, who was a professor at the University of Glasgow, and Charles (Chuck) Thacker who was in at the start of two famous Silicon Valley labs: Xerox Palo Alto Research Center and Digital Systems Research Center.

Needham was a consultant for both. The Digital centre, "one of the best research labs in the world", is his model, though Xerox Parc was run on similar lines.

"We do not want to grow beyond about 50 because Chuck and I know that up to that number you can have a flat management structure. That is very advantageous for research because it means that people can move interests without moving about the organisation, because there isn't an organisation. That succeeded spectacularly at Xerox Parc. It left Chuck and me with the solid conviction that that is the right way to do it."

They could not do it at Redmond, Needham adds. The place is too big. But in Cambridge he expects no problems. "The big challenge, I think, is not that of running a first-class lab. It may seem arrogant but I am totally confident we can do it. I have done it before. Chuck has done it twice. Mac has done it once. The challenge is being eight time zones away from head office and eight time zones away from the people who have got to be aware of us and pick up our ideas and use them for something."

The airlines are going to do good business. "Although email is truly wonderful and Microsoft is the most email-based company you could possibly imagine, you do actually have to know the people, you have to know who would be likely to be interested in something."

Staff joining the Cambridge lab are usually sent off to Redmond after their first month, just to get to know people. "We are not seeking to get them indoctrinated," Needham insists. "Our culture is our own."