Roughly 7 or 8 months ago a rather interesting discussion appeared on the ol' Backtrack 4 forums on brute forcing. You can view that topic here, http://bit.ly/nB2diq. The author, CKing, is mostly (to my understanding) just conceptualizing a free-to-use online WPA brute forcer for Backtrack Forum members. Well, he mentions that cost would still be an issue, but that he'd fork up some preliminary hardware. This got me thinking... why not build a WPA Brute Forcing (or even brute forcing in general) distributed computing project. I think it'd be best not to focus on WPA only (especially since there's already a fatal flaw in TKIP I've been looking into... those of you wondering why tkiptun-ng doesn't work, I can help), it'd be especially nice if we could make a versatile platform that can run a number of brute force attacks.
It would be much like the folding@home (http://bit.ly/9yJHvq for more) or other @home distributed projects. Distributed computing harnesses millions of volunteer computers as a parallel supercomputer. This means we'd need a willing audience. To help with this I'm thinking the client used to perform an attack would require that you've already donated somethign like 24 hours of GPU/CPU time. Lots of people with small servers could easily add a VM to handle additional volunteerin which would be nice. Someone may have to run a centralized server, which I'd be willing to sink a little "one-time-only" money into.
As far as the programming goes, I'm not too familiar with this sort of project, but Berkley offers some tools like BOINC (the Berkeley Open Infrastructure for Network Computing. More here, http://bit.ly/aUbL0P). Clients would be supported on all OSs.
That's about as far as I've gotten, right now I'm really just looking for reactions and maybe a few dedicated people with some basic (or expertise) programming skills to contact me. You can email me directly @ email@example.com
P.S. The project would be called TRANSLTR.