That PlayStation 3 you may find under the tree this holiday season is more than the hottest gaming console around; it's also a powerful tool in fighting protein-based diseases. Using distributed computing technology, Sony has teamed up with Stanford University's Folding@home project to take a quantum leap forward in simulation processing power.
Proteins, the building blocks of the human body, must take shape or "fold" before they can carry out their functions. When they don't fold correctly, scientists believe, they can lead to disorders like Alzheimer's disease, Huntington's disease, Parkinson's disease, and certain forms of cancer. But a single computer would need 30 years to simulate a single fold.
Founded in October 2000, the Folding@home project links about 1.8 million CPUs around the world to run these simulations. So far, it has unlocked the processes behind several small, fast-folding proteins. The Cell processor in the PlayStation 3 will help the cause immensely, as researchers expect performance akin to 100 GFLOPS per computer.
If 10,000 PlayStation 3 consoles join the program, the researchers say, the program will be able to achieve performance on the petaflop scale. Considering the console's popularity, this shouldn't be a difficult figure to achieve. Owners simply install and enable the Cure@PLAYSTATION3 client, which connects to the Folding@home network when the console is on but not being used.
In addition to its processing power, the PlayStation 3 offers another advantage. Its advanced user interface uses the console's RSX graphics engine to visualize the simulation process. Users can view the target protein as it "folds" in a 3D space (see the Figure). They also can rotate, zoom, and move the image in real time.
Sony and Stanford are now testing and further optimizing the Cure@PLAYSTATION3 client. Meanwhile, researchers believe the processes behind protein folding could be applied to the design of protein-sized nanomachines with their own biomedical applications, including finding a cure for these diseases.
Reprints
