Online network for solar research

Distributed computing is a valuable tool for scientists with data to analyse

A joint project between IBM and Harvard University is tapping into thousands of computers, in an attempt to discover new solar and energy storage materials.

The Clean Energy Project uses distributed computing to number-crunch molecular equations.

Users volunteer the processing power of their computers, with a grid server assigning tasks.

The project is part of IBM's World Community Grid, which assists research projects that "benefit humanity".

Other efforts include researching new cancer-fighting drugs, improving the nutritional content of rice by examining new protein structures, and trying to prevent the onset of AIDS by blocking HIV protease, the enzyme that makes HIV infectious.

The Clean Energy Project has two goals: finding new organic photovoltaics to provide inexpensive solar cells, and then researching possible polymers for the membranes used in fuel cells.

Heavy in light

Traditional solar cells, made with silicon, are usually heavy and inflexible. Organic solar cells have the potential to be much cheaper, flexible, and lightweight.

However they are not as efficient when compared to traditional cells, and their working life is shorter.

The lead researcher behind the project - Professor Aspuru-Guzik - said that currently solar cells using organic molecules have an efficiency of between 5 and 6%, but that an efficiency of 15% was required to make them commercially viable.

The task is to perform chemical calculations on tens of thousands of organic molecules to determine which show the most promise, utilising the CHARMM molecular mechanics package developed by the Karplus group at Harvard University.

Other grid projects include unravelling and identifying human protein chains

"It would take us about 100 days of computational time to screen each of the thousands of compounds for electronic properties without the power of World Community Grid," he said.

By distributing the computations among thousands of volunteered PCs, the project could be completed in two years. By contrast, a traditional supercomputer approach would take more than 20 years.

Distribute and conquer

Modern home computers can perform billions of calculations a second. Most of the time, that's far more power than the average user needs - so even though you're working hard, your computer is just ticking over. Distributed computing projects make use of this spare computer potential.

Distributed computing is a particularly valuable tool for scientists who have large amounts of data to analyse, or who are modelling very complex systems.

For example, the Folding at Home project, which aims to unravel molecular protein chains is now in its eighth year, and there are now more than 3m computer users working on the SETI@home project looking for signs of contact from alien life.

IBM hope that their World Community Grid can attract a similar number of users.