As computers get smaller and faster, scientists have started to explore ways of reducing the size of physical components to the level of single atoms and molecules.

The design and manufacture of these microscopic structures is called nano-technology.

Over the past few decades, the silicon chips which form the heart of computer systems have become more complex.

Silicon transistors are getting smaller, so an ever-increasing number can be packed on to the thin slice of silicon that makes up the chip.

Making transistors smaller makes them switch quicker, which is why this year's computers are invariably faster than last year's.

Now scientists and engineers at IBM can foresee the end of silicon miniaturisation of this kind and predict a new kind of biologically-developed processing device, called a carbon nanotube transistor, containing far fewer atoms.

IBM scientists have been working with carbon nanotubes for several years.

They know they can be made smaller than the smallest-possible silicon transistor but are not quite sure if they will also be faster.

So far, things look very promising.

If this, or some other new nanodevice re-places the silicon transistor, the cost of computing power will continue to fall for decades to come.

But nanotubes come with their own set of problems. When we build a computer chip, designers specify the position and dimensions of every part of every wire and transistor on the chip.

But when we build things as small as single atoms and molecules, chemistry de-fines the end-result.

Carbon nanotubes are built by chemical synthesis. We don't have to specify the position of every atom. We just have to provide the right catalyst and reaction conditions and the tube will grow itself.

Unfortunately, we are not yet very good at this - not good enough to manufacture the hundreds of millions or billions of identical and precisely-positioned nanotubes required for the manufacture of future computer chips. But we are working on it.

Nanotechnology goes a lot further than just transistors.

Imagine nanostorage, where more than 100,000 DVD movies could be stored inside your wristwatch.

This is not just a dream because scientists at IBM have already developed a new type of storage called Millipede that has more than 1,000 tiny arms which can all write data simultaneously on to a microsc-opically-thin film of polymer.

IBM is confident its research is the future of computing but said most of the work in its labs was not expected to yield results for between five to 15 years.

Physical sciences director Tom Theis said: "We're bringing together researchers in physics, chemistry, materials science, electrical engineering and biology. We are only beginning to understand how we might be able to make useful devices at the atomic scale."

Many new, nano-structured materials are already in commercial production and users can expect to see new nanosensors for detecting chemical and biological hazards in the near future.

IBM's Millipede info-rmation storage device may enter the market in a few years.

So, will nanotechnology revolutionise the computer industry? IBM thinks so and has instigated a major nanotechnology research effort to explore ways it can utilise the new developments.

We can make predictions of how nanotechnology can help im-prove computer chips and storage technologies. There are many ways it may impact upon us that are, as yet, unimaginable.

Take the case of the ENIAC supercomputer. Built in 1946 by researchers at the University of Pennsylvania, it took up several floors of a large building and was the fastest computer in the world.

It appeared on the cover of Time magazine with the headline Giant Brains.

Today, you can find the processing power of the ENIAC in one of those greeting cards that play a song.

In the same way, we cannot predict the full impact nanotechnology will have on our future.

www.ibm.com