Folding@home Petaflop Initiative (FPI)

Introduction

Since 2000, Folding@home (FAH) has allowed a major jump in the capabilities of molecular simulation. By joining together hundreds of thousands of PCs throughout the world, calculations which were previously considered impossible have now become routine. FAH has targeted the study of protein folding and protein folding disease, and numerous scientific advances have come from the project.

Now in 2006, the FAH team (with key collaborators in industry and at Stanford University) has developed new computational methods that will allow for another major advance in capabilities. This advance utilizes streaming processors now common in inexpensive consumer electronics, such as the Cell processor in Sony’s PlayStation 3 or personal computers with Graphics Processing Units (GPUs) from ATI, to achieve performance previously only possible on supercomputers. With this new technology, we are able to attain performance on the 100 gigaflop scale per computer, at a very modest cost (~$500). We are beta testing the ATI GPU client software internally at the moment and will likely announce an open beta in four to five weeks (at the end of September).

Thus, armed with this new technology, we are setting out on a new initiative to take Folding@home to even greater heights. By combining merely ~25,000 computers (each with some sort of streaming processor), we could perform calculations on the Petaflop scale (1,000,000,000,000,000 floating point operations per second) – a level of performance currently unmatched even by the fastest supercomputers. As Folding@home currently consists of approximately 200,000 actively processing computers, we expect that as this hardware becomes more common, we would easily surpass the 10 Petaflop level.

Our goal is to apply this new technology to push Folding@home into a new level of capabilities, applying our simulations to further study of protein folding and related diseases, including Alzheimer’s Disease, Huntington's Disease, and certain forms of cancer. With these computational advances, coupled with new simulation methodologies to harness the new techniques, we will be able to address questions previously considered impossible to tackle computationally, and make even greater impacts on our knowledge of folding and folding related diseases.

For more details see the Folding@home web site or contact Professor Vijay Pande. Folding@home is supported primarily from grants from the National Institutes of Health and the National Science Foundation.

Frequently Asked Questions

Why push for new technologies, such as GPUs or the PS3, rather than try to recruit more CPU clients?

In order to tackle many of the problems of interest (especially related to protein misfolding and aggregation, such as in Alzheimer's Disease), we need to not just have lots of computers participating, but we need results returned more quickly so that we can simulate trajectories of sufficient length. Right now, we achieve this by running for many months to years (indeed, our first Alzheimer's Disease simulations ran for almost two years straight). That's where the new clients come in. They give us considerably longer trajectories in the same wall clock time, allowing us to turn what used to take years to simulate even on FAH, to a few weeks to months.

What types of calculations do the new clients improve speed?

There are primarily two types of calculations we perform, which differ by how we simulate water. The GPU and PS3 clients greatly speed "implicit solvation" simulations, in which water is handled mathematically in a continuum fashion (see the Wikipedia article on implicit solvation for more information). Our SMP client (discussed in our High Performance FAQ and SMP FAQ) will significantly speed "explicit solvent" simulations, where water atoms are handled atom by atom, in an explicit fashion, just like any other atom in our system. Currently, the GPU & PS3 only significantly speed implicit solvation and the SMP client only speeds explicit solvent, so each has its limits, but together they work to give FAH considerably more computational power than ever before.

For More Information, Please See:

• FAH FAQ
• High Performance FAQ
• SMP FAQ
• PS3 FAQ
• Folding Community Forum

Last Updated on January 24, 2008, at 06:14 AM