utilizing GPU's for audio processing

[lung]

I just came across this interesting posting over at Slashdot which talks about using the incredible processing power of the GPU for audio processing. There is a link to a paper on the subject, as well as some other articles on using GPU's for FFT transformations. There is a lot of power to harness, and in the Slashdot discussion mentions that the UAD-1 "had a previous life as a video card and a DVD hardware accelerator".

I thought this might interest some of you, especially those of you developing resynthesis plugins or other high CPU load plugins.
Cheers