voodooattack Lv 1
Would it be possible for an OpenCL/CUDA/DirectCompute powered build of foldit?
Right now it's very CPU intensive and a parallel computation back-end would surely help the lag.
Would it be possible for an OpenCL/CUDA/DirectCompute powered build of foldit?
Right now it's very CPU intensive and a parallel computation back-end would surely help the lag.
I second this idea - access to more cores always means more speed.
Bump.
Maybe contact to Stanford and borrow some Folding@home GPU code?
Hell yea! Take it! :)
Duplicate: http://fold.it/portal/node/990207