Thirty yards away, a modest roundabout with a small green sign finally tells you where you are. Welcome to Cambridge Science Park, 130 acres of research courtesy of Cambridge University's Trinity College, the landlord, and 4,500 scientists, suits and sandwich ladies.
With over 70 firms, the park is the largest concentration of information technology and drug development companies in the country and one of the biggest in Europe. It is also a model "cluster", the government's new- found instrument to inject efficiency and entrepreneurial spirit into UK plc. According to "clustering" theories, the physical proximity of hi-tech businesses at an early stage of the development of their products encourages collaboration and helps them survive the difficult start-up years.
The government is so keen on clusters that the Trade and Industry Secretary, Peter Mandelson, is set to make them a key part of New Labour's industrial policy in a white paper on competitiveness to be published next month. Mr Mandelson is said to have been very impressed by the idea during a recent visit to Silicon Valley and plans to trigger a "cluster explosion" in the UK over the coming years.
Scottish Enterprise, the development agency, has been among the first to move, by appointing Bob Downes, one of its top officials, to the brand- new post of "head of clusters".
But do clusters such as Cambridge Science Park work? And can they live up to the government's dream of being efficiency-enhancing islands of research and development?
Most of the tenants of Cambridge Science Park do not buy the government's rhetoric. John Padfield, the chief executive of Chiroscience, one of the UK's largest and most successful biotechnology companies, is a classic example of a dissatisfied cluster customer.
Dr Padfield's company has spent more than six years in the Cambridge cluster and is now moving to a different site south of the city because it feels the park has failed to deliver on its promises. His complaints highlight one of the key issues for present and future clusters.
If business groupings are to succeed, physical closeness, is not enough. To spur rival firms into interacting business parks must foster a "sense of community" which encourages people to share ideas.
"The idea of clusters is an excellent idea in theory, but if someone asked me: `Has Cambridge Science Park created a community or is it just a physical collection of people? I would have to reply the latter".
In the experience of Dr Padfield and many other executives, the famed interaction with other firms has been close to nothing and the benefits of pooling resources and bouncing ideas off of each other is almost non- existent.
Part of the problem is in the lack of common facilities where the mingling between the cluster's inhabitants can take place. The lesson that future developments will have to learn from Cambridge is that scientists need to be dragged away from their laboratories by the lure of common amenities. In the Cambridge Park these are conspicuous by their absence.
With its low-rise buildings nestling among hills and lakes, the park looks more like a elderly people's holiday camp than a thriving centre of business development. Walking around the park at lunchtime is a bit like taking a stroll in the Sahara. There is only one restaurant and conference hall, the Trinity Centre, which is regarded as totally inadequate by most tenants. One executivesaid he would be "embarrassed" to bring his customers there. The companies have complained to Trinity and the college has launched a major multimillion pound plan to build a new conference centre and health club by the Millennium.
This should drag the boffins out of their shells and should help fulfill the park's potential, according to its supporters. They point to a number of other advantages brought about by clustering. First, the use of common suppliers for things such as technical equipment and building maintenance triggered sizeable cost-savings for the park's tenants. More importantly, cluster enthusiasts believe that having a hi-tech grouping with close geographical and financial links to a hotbed of research such as Cambridge University is a major drive in staff recruitment. Stephen Inglis, the research director at Cantab Pharmaceuticals, another leading biotech firm, says that the eight years spent at the park "have been great".
"There is a warm glow associated with a company which is linked to Cambridge and is next to the University. Being here gave us the credibility to attract the right-quality people". John Brown, the chief executive of Peptide Therapeutics, another drug company, agrees.
"Cambridge is a great attraction for the kind of people we want to employ. They like to work here because they like the hi-tech culture".
Dr Inglis believes that, although Cambridge Park has been going for almost 30 years, it needs more time to bear fruit. "There is a tremendous wealth of knowledge and this coalescence of interest will create its own momentum."
Being able to attract the country's best brains is a powerful selling point. But if the Cambridge experience is anything to go by, even that advantage could be jeopardised by poor infrastructure.
Colin Webb, the head of European development at the US biotech giant Amgen, said that headhunters told him that potential applicants for one of the company's positions had been put off by the park's location. Traffic is one of the location's biggest drawbacks.
The consensus among the park's workers is that peak-hours congestion, when most of them get to work, is horrendous and makes entering and exiting the park a harrowing experience. One worker said that it takes him 45 minutes, on average, to leave the site every evening - hardly the kind of lifestyle to endeavour the park to the UK's best scientists.
Staff can also be deterred by high house prices in the city, which are partly due to the presence of the park.
As one executive put it: "Nobody knows whether clusters work, but the clear lesson from Cambridge is that they certainly do not work when the infrastructure is not right."