Folding has thrown a lot of programmers at GPU computing and rewritten some programs to concentrate on the things that can be done quickly on such processors. (The CPU is still more general purpose so some functions were cut - though I have not heard if this was done for the PS3, which may be more powerful than most GPUs.)

So it makes a lot of sense to go with Folding. We are investigating the PS3, but we have our hands full with new projects right now.

Lawrence

F@H started a Windows SMP beta on March 18.....their bets are not on 1 horse.... that's the difference with a research driven project.

check this out

freakin amazing


This is freaking amazing. I was checking out some message boards last night at the Folding Forums at Stanford, a group that tracks the Folding at Home application. You know, that's the software that runs on Sony PS3 or PS2 gaming consoles, all linking up over the Internet and using their spare cycles to help the university process vast amounts of Alzheimers research data? To my astonishment, I discovered that a small legion of 13,000 PS3s running the Folding at Home app account for most of the computing power in the project, amounting to about 56 percent (PS3s = 316 measured TFLOPS) of the total.

On average, between the superfast and superslow PCs, 159,033 PCs are only doing about half that much. (151 TFLOPS). Essentially, 13,000 PS3s have just made the Folding at Home Project the fastest distributed computing project on the planet, ever. (I believe it used to be SETI @ Home, which was something like 280 TFLOPS.) This also means the PS3 met Stanford professor Vijay Pande's expected one-month goal in one day. (We'll update this post with confirmation once Dr. Pande gets back to us.
http://gizmodo.com/gadgets/home-entertainment...spitting-range-246664.php

Same applies to the gpu client, the windows smp client and the linux smp client.

Hats off though to a progressive project