Future developments in computer technology will increasingly be driven by the search to understand life itself.

In the past, the primary drivers for high performance computing were large state-funded projects such as weapons testing, space exploration and the charting of global weather.

Many of these areas still push the boundaries of technology, but the biggest challenge demanding increasingly fast supercomputing power is life - from helping to improve the discovery of new drugs to the modelling of biological processes such as protein folding.

IBM is at the forefront of this trend, with researchers in laboratories across the world working on a diverse range of projects, such as the modelling of proteins and enzymes.

'If you think about biology in its most abstract sense, in terms of what happens in living things, it consists of coding of information that is in the DNA and the processing and storage of that information,' said Dr Paul Seidler, manager of science and technology at the IBM Zurich research laboratory.

'That sounds like a description of the IT business, not of biology. If you think about it that way, it's almost inevitable that IBM gets involved in life sciences.'

At the root of IBM's life sciences research is its virtual computational biology centre, where researchers explore what could be achieved in bioinformatics - how computers can help biologists do more.

'We have work going on to really push the limits of computational methods and try to do the most accurate and precise simulations of biological molecules,' said Seidler.

At the Zurich laboratory, researchers have been simulating how progesterone - the 'pregnancy hormone' - binds to a specific protein in the human body.

Understanding this interaction is crucial for the development of pregnancy-related drugs.

Last month, IBM announced a major step forward in its research, unveiling a methodology that sheds light on how progesterone binds with its human receptor.

The breakthrough may reveal critical information for the development of progesterone-based pharmaceuticals, as well as demonstrating how virtual experimentation can potentially reduce the time and expense of bringing new drugs to market.

The crucial difference from past techniques is that this research is based on 'in-silico' methods, using computers to calculate the results, rather than working with living cells.

Automotive and aeroplane manufacturers have been using computers for decades to cut the costs of developing new products, but this remains an embryonic tool for the biological industry.

Key to achieving this is massive computing power, which is used to calculate the complex mathematical algorithms required. To facilitate this, IBM installed what was then Switzerland's fastest supercomputer at the Zurich lab in 2002.

'These calculations require sophisticated software algorithms and massive parallel computing power because of the huge amount of data and number of process steps to be computed for a reliable simulation of any real system,' said Wanda Andreoni, IBM's manager of computational biochemistry and materials science at Zurich.

'We want to be useful to pharmaceutical companies, and at the same time we want real problems to work with, which stimulate us to develop new methods to satisfy the requirements of this field,' she said.

'Previous methods didn't deliver sufficient results to make this worthwhile for pharmaceutical firms. This is now starting to change.'

Today, life sciences-related research encompasses a diverse range of technologies, including grid computing, visualisation, data and knowledge management and micro-fabrication processes.

Much of the supplier's research efforts have not resulted in actual products or services for it to sell, but some results do drive new product development.

The Blue Gene/L project, which was created to tackle the challenge of protein folding, intends to deliver an order of magnitude change in supercomputing power.

And while life science is the driver, the project's applications will be felt in a range of more traditional markets for IBM, from financial services to weather prediction.

'One goal is to advance the art of biological simulations, but in general it is to advance the art of computer design,' said Seidler.

Much of the work also translates into significant and ongoing cost reductions in products.

For example, the Zurich lab recently implemented a new server based on IBM's latest commercial technology, which is as powerful as one of the supercomputer racks installed in 2002.

Back then, one of the racks set the company back a cool CHF4m (£1.75m) - today's server cost just CHF100,000 (£44,000).

These results help to drive IBM's business forward, but the supplier may deliver important benefits along the way, such as helping to improve fertility or contributing to breast cancer research.

The future of computing?

Last week, two prototypes from IBM's $100m (£55m) Blue Gene research project were ranked among the top 10 of the world's most powerful computers, according to the TOP500 supercomputing list.

The number four-ranked Blue Gene/L DD1 Prototype, which provides a peak speed of 16 teraflops from its 8,000 PowerPC processors, is about the size of four large refrigerators.

But it delivers only one sixteenth of the planned processing power of the full Blue Gene/L system, which is due to be installed at the Lawrence Livermore National Laboratory in California next year.

The world's fastest supercomputer, the NEC Earth Simulator in Japan, occupies the space of a football field. When it's built, Blue Gene/L will deliver eight times the processing power, in the space of a tennis court.