In the early days of computers, it was realised that electronic technology would be orders of magnitude faster than mechanical devices. Put simply, when a human operator turns a handle to add together two numbers, an early electronic computer could do the same in a matter of one thousand of a second – and today’s machines can do this in better than a million and more times a second. This increase in speed enabled research scientists to use a computer to solve problems which were previously considered not possible. Electronic technology increased the scope of research generally in all subjects, scientific and non-scientific.

By the end of the 1950s virtually every institution had at least one computer, and, thanks to enlightened government support, funds were provided for the necessary staff and expertise to establish Computing Services.

Things have changed radically since the 1950s but it took a further 30 years before the revolution of optical fibres. This new technology created plastic cables containing glass fibres through which a beam of light was carried and at an enormously greater capacity than electrical pulses through metal cables.

In 1970 computing was already spreading as an essential tool in many subjects and beginning to transform academic research. My colleagues in the UCS had to keep up with quickly advancing technology which, fortunately, coincided with a period of generous government funding to employ the necessary additional staff. Other benefits that came from continuing technological advances were that computers became considerably more powerful and didn’t break down so often.

The Collegiate system at Cambridge University is complex. By and large every student belongs to a single College where they live and receive tutorial supervision. Each is associated with one or more Faculties where they receive formal and practical class teaching and undertake research. This is a complex arrangement, and a similar structure applies to academic staff.

Starting with machines of the earlier days, computers began to be connected to terminals and other specialised machines. Machines on a single site could be connected by our own cables simply by installing a cable from one building to another. But there was a law that required any such connection that crossed a public highway had to be provided by, and rented from, the national telephone company. The cost for any such connection was a fixed monthly fee dependent on distance and independent of the amount of traffic.

In the early days, the buildings of the average university were typically located on a single campus, most of them large areas with few, if any, public highways. But Cambridge University is spread around numerous sites throughout Cambridge City: some sites are quite large, others are isolated buildings. With public highways under control of the national phone company, this was generally considered too expensive. and just impractical. This layout of Cambridge city may be fine for students and tourists, but impractical for us when connecting many users to a central service.

An early solution was to design and develop a so-called PAD (Packet Assembler Disassembler) which was a small electronic device that multiplexed up to 16 devices linked by a single public line to a central switch in the UCS main computer room. We arranged for a number of these devices to be designed and manufactured by a company known as CAMTEC (unrelated to the University). Being made of microelectronic components, a PAD was a low-cost way of connecting several terminals in a one building through a single link to another. In this way we had a solution to our networking problem.

Nevertheless, there was a steady increase in the number of terminals, some with printing devices others with video displays. These were deployed in numbers across the University, each connected to a nearby PAD which in turn was connected to a rented communications link. PADs could also link to facilities in other departments, enabling the sharing of information across the University. The number of connections grew by which time the central computer room had become a hub for real networking.

The demands for growth were steady but there was still a major problem. PADs were limited in various respects. They were suitable for terminal to computer connections but not for high-speed connections between two computers. If a remote connection was across a single site, we could and did, lay our own cables from building to building, but otherwise circuits had to be obtained from the monopoly public supplier.

By the 1980’s there were two solutions on the horizon: one technological – the other political. Instead of copper cables and electric current, industry was developing fibre-optics using light passing through glass fibres. Glass is significantly cheaper than copper and light is very fast indeed. This new technology enabled digital transmissions at higher bandwidth (and over longer distances) by orders of magnitude.

The necessary political solution started to appear when industry began to demand government treat seriously the monopoly protection of the telephone companies. The need was for education and research to be allowed to pay for their own cables to cross public highways between its own properties. By the early 1990s, we began to see fundamental changes in government thinking about public communications.

The GBN today is a broadband network, using modern fibre-optic cables carrying information at the speed of light. Usually deployed in underground ducting, the network is low maintenance and consumes little power – in brief, it is green, inexpensive and relatively future proof. The network was built and became operational in 1992 building on technological advances some made in Cambridge since World War 2. No doubt, more revolutionary developments will come in future, and the GBN, or its successor, will continue to serve research, teaching and administration until something new comes along and provides even better ways of doing useful things.

Finally, a personal remark. In 1994 after 24 years, I left the Computing Service for new pastures. The same Vice-Chancellor, Sir David Williams, in his annual report to the University graciously acknowledged me with the words

My colleagues from the University Computing Service particularly appreciated Sir David’s use of the words ‘courage and imagination’.