Folding@home
Folding@home is a distributed computing project that uses a series of interconnected computers that have downloaded and installed the program to help simulate and study protein folding to find possible protein combinations that may cure common diseases such as Alzheimer's and Parkinson's.
The Folding@home client runs in the backround of these personal computers and other devices, taking advantage of unused CPU power. It periodically connects to a server to retrieve packets of data or "work units" to perform calculations, processing protein structures while the computer is idle and then sent back to the server when done.
The Folding@Home client is free to download from the project's website. The website keeps tracks of statistics, including the number of active CPUs across a variety of platforms. Windows and Mac users (using both PowerPC and Intel processors) can download and install the client on their systems and run it whenever they see fit (or set to automatically run in the background).
Folding@Home also takes advantage of high-performance graphics processing. The potential of general purpose processing using GPU cores is becoming more a part of every day life on the web, accelerating resource-demanding applications and multimedia tasks. Folding@Home has special modified clients that use the overwhelming number-crunching capabilities of high-performance GPUs from both ATI and NVIDIA. The site's own stats show a remarkable performance advantage when the GPUs are in use, compared to standard CPUs in Windows and MAC computers.
Folding@Home has also jumped into living room consumer electronics products through integration with the PlayStation 3 (PS3) console. Using the Cell processor, PS3 owners connected to the Internet can take part in the Folding@Home project too, and at the time of writing (March 2011), the Folding@Home website records 27,665 active CPUs from PS3 consoles (1,087,723 total PS3 CPUs). The website defines an active CPU as one that has returned Work Units within the last 50 days (10 days for GPUs).
