There are no flashing green lights or sounds that go "bing", no sweetly sinister disembodied voice: just a passageway lined with grey boxes. Meet Hector, weather expert, inventor and prophet, resident of Midlothian - and Britain's most powerful supercomputer.

Hector, which stands for High-End Computing Terascale Resource, will be officially launched on Monday by the Chancellor, Alistair Darling, at Edinburgh University's Bush House site, south of the capital. The £113m service, at the university's advanced computing facility, will run for six years and be operated by EPCC (Edinburgh Parallel Computing Centre). One of the top 20 most powerful computers in the world, and the first and largest of its type (a Cray XT4) in Europe, Hector is a coup for Scotland and for Edinburgh University in particular.

What sets Hector apart is its size. Because modern technology allows immense processing power to be condensed into small spaces, a vast amount can be packed into a large computer, and Hector is a giant. It covers 1800 sqft of floor space, fills 60 wardrobe-sized cabinets and comprises 11,000 processors with the combined power of 14,000 high-spec PCs. With that size comes mind-boggling speed.

Professor Arthur Trew is service director at EPCC and the man responsible for making sure Hector runs smoothly. He puts it this way: "A laptop will do a thousand million calculations a second. Hector will do 50 million million calculations per second. Hector does 10,000 calculations per second for every person on the planet."

Phew. And the capability it brings is changing the whole nature of scientific inquiry. Science is advanced by testing theories with experiments, but what supercomputing allows is the creation of sophisticated models and simulations through which hypotheses can really be put to the test.

The construction of Hector was funded and managed by the Engineering and Physical Sciences Research Council (EPSRC) on behalf of the UK Research Councils. Jane Nicholson, head of high-end computing at EPSRC, says: "Hector will enable us to do research that we simply could not do in any other way. It is needed in areas such as the latest climate models and for the design of aerofoils, which have a direct and important relevance to society.

But it is also important for fundamental, curiosity-driven research. We want to push forward the boundaries of knowledge and we need to be sure that the UK research community can carry out leading-edge work and be able to compete internationally."

Professor Trew adds: "Hector is the latest in a series of national high-performance computing facilities, the newest, biggest and most expensive we have had.

Hector came about owing to the demand for these big computers to solve research problems."

These are very wide-ranging, from designing new drugs and producing more detailed medical images to examining the distribution of fish eggs by ocean currents. But one of the most significant is predicting the future global climate. Existing computerised climate models create simulations of global weather patterns, but with relatively limited levels of detail. According to Dr Len Shaffrey of the National Centre for Atmospheric Science, a typical model maps the Earth using a grid which divides the atmosphere into 300 sqkm boxes and the ocean into 100 sqkm boxes. "However, at this resolution many processes, such as the impact of mountains on the atmosphere and small-scale eddies in the ocean, are not properly represented," he says.

Now researchers want to use Hector to produce a more detailed picture, dividing the atmosphere into much smaller 60 sqkm segments and the ocean into 30 sqkm patches. Perhaps this higher degree of detail will improve the accuracy of prediction.

For Richard Kenway, Tait Professor of mathematical physics at Edinburgh, Hector has a double significance. On the one hand, Kenway is vice-principal for Computational Science, with overall responsibility for Hector; on the other, he is enthusiastic about what Hector might be able to reveal in his own research area.

According to Professor Kenway, physicists already have "a very good understanding of the forces of nature", or at any rate, of three out of the four forces that govern the universe. They have a model of how nature works through these three forces, called the standard model.

But, as yet, there is no theory that encompasses the fourth force, gravity. "One of the holy grails of physics is to discover a single theory of everything, which describes these four forces and how all the elementary particles interact with one another through them," says Professor Kenway.

Physicists have a big collective hunch, though, that there's a piece of the jigsaw missing. "We don't know what it is," he says. "We are missing something, a piece of physics that we haven't managed to discover in all the experiments."

The aim with Hector is "to figure out what this new physics is". Kenway and his colleagues will do this by running the standard model on Hector to get a simulated set of results and then compare them to the results of real bona-fide experiments in a particle accelerator. Because Hector is so powerful, the results they get from their simulation should be very precise. Any discrepancy between them and the actual experimental results, then, "really can be attributed to something new, to a new discovery".

All this will take a lot of hard work, Professor Kenway admits, but the stakes are high. He says: "It is exciting or we wouldn't do it. This quest is something that physicists around the world are engaged in. When we do our calculations on Hector we are combining the results with calculations on big computers that our colleagues in the US are using. Hector is very powerful, but in order to do the sorts of things I describe we need several machines like that and we collaborate.

"Hector puts us at the forefront of the field. In some sense, it is our entry ticket to play the game. Hector is powerful enough at the moment to be competitive with anything that the Americans have."

How Scotland's weather will look in 100 years' time and the laws that govern the universe are quite a lot to be getting on with, you might think, especially as Hector is not going to be around for that long. "Like a laptop, after a few years you replace it," says Professor Trew. "We are planning for Hector to have two upgrades - the upgrade in two years will make it five times more powerful and there will be another in four years to improve the performance even further."

But, in spite of its short six-year lifespan, universities all over the UK are lining up to use Hector and, indeed, some research has already been done. Scientists from Southampton University, for instance, have been using Hector to model the behaviour of fluid flows, which could be useful in understanding air flow over aircraft wings. Turbulence is particularly difficult to simulate because it involves eddies with a wide range of length and time scales. Using Hector, the team have constructed the largest fluid simulation ever carried out in the UK.

Researchers can get into Hector over the internet. "No-one comes to visit, it's not that exciting once you get here," says Professor Trew.

But while it can be accessed remotely, that doesn't mean that just anyone can use it. If you want to use Hector, your research proposal has to merit it. "Because Hector is the biggest and fastest, you have to be very careful to use it most effectively," he says. "The process of using Hector for a project is one of peer review. You must write a scientific justification for the work." A user name and password is required and "unless you are a registered user then the firewalls will stop you getting close to it".

Supercomputers are not cheap to run, after all. They use a lot of energy and get very hot - which is why Scotland is a prime location for them. "We are putting in the minimum amount of energy needed to keep the machine going and are relying on the Scottish weather to keep it cool," says Professor Trew. "While it uses a fair bit of energy, we have given a lot of thought to how we minimise that." Even so, with six years of continuous use ahead, he's in for one scary electricity bill come 2014.

rebecca.mcquillan@theherald.co.uk

Mother of all motherboards

  • Hector is a Cray XT4 and computes at a rate of over 50 teraflops (50 million million calculations per second).
  • Colossus, the code-breaking machine used during the Second World War, is sometimes called an early supercomputer, but experts regard IBM's Naval Ordnance Research Calculator in the 1950s as the first - it was used to calculate missile trajectories.
  • The father of supercomputers is Seymour Cray, who worked for Control Data Corporation in the 1960s, creating the CDC 6600, before leaving to start his own company, Cray Research, in 1972. The European Centre for Medium Range Weather Forecasts took 12 days to produce a 10-day forecast in the late 1970s using a CDC6600. They later switched to a Cray 1-A which produced the same forecast in five hours.
  • As computers have become more powerful, they have got smaller. A single modern desktop PC is now more powerful than a 15-year-old supercomputer.
  • The fastest computer in the world is the IBM BlueGene/L System at Lawrence Livermore National Laboratory in California, which researches biological processes. The current system, a cluster of 65,536 computers each containing two processors, can perform 478.2 teraflops.
  • The most famous supercomputers of all, though, are fictional, including Arthur C Clarke's HAL 9000, the paranoid, murderous supercomputer in 2001: A Space Odyssey. Sardonic but benign by comparison is Red Dwarf's Holly, while Douglas Adams's Deep Thought took 7.5m years to conclude that the meaning of life was 42. In the real world, IBM Deep Blue, the machine that beat chess champion Garry Kasparov, became a celebrity of sorts.