Distributed Computing

As seen around the web over the last couple hours, Folding@home has released a client to model proteins using your ATI graphics processor (X1900). Unfortunately, it seems like its only a very early release, requiring a specific driver version and sub-version of directx 9.0c. This is a great idea though, whatever the implementation problems, since the GPU is a very powerful math processor that sits largely unused unless you are playing games.  AnandTech has a good summary of the Folding@home project and the new GPU project if you want to read more, but basically the idea is that the more complex proteins (very long chains of peptides) bunch up or "fold" into very complex shapes due to the interaction of the atoms ( hydrogen bonds, etc). If you put the protein under an environmental stress (change its temperature, ph, etc) the interaction of these forces will change and the protein will "unfold". Depending on the protein, it may or may not "fold" again when its environment returns to normal (some may also require interaction with other molecules to fold). Since many (most?) of the human body's functions require on proteins (particularly enzymes) the hope is that with some study and modeling scientists will be able to figure out exactly how some common proteins work and prevent or cure problems.

Anyway, enough bio,  if you want to help out and have an ATI X9100 take a look at the ATI FAQ here.If you don't have an ATI video board (NVidia has better Linux support), you can get a client for your CPU here (be sure to run two copies if you have a dual core processor). Once running, you shouldn't really notice the client since it runs at a low processor priority.

If you don't like Folding@home, there are plenty of other options if you are looking to put your CPU to work, among the more well known ones:

SETI@home (find aliens)
ChessBrain
Electric Sheep (Screensaver Generator)
GIMPS (Great Internet Mersenne Prime Search)
University of Oxford "Screensaver Lifesaver"

Some more can be found on this list

Happy Computing,

(and feel free to correct me if I've forgotten too much Bio already)