Alby_da_Wolf on 27 January 2014
@those asking why using the GPU for audio:
I haven't read the whole thread, but just in case nobody wrote it yet, yes, it makes sense. A GPU has HW features and optimisations that allow it to do DSP work a lot more efficiently than a general purpose CPU. It routinely does it when it applies filters to graphics data, for example, and add-multiply operations are also routinely used when multiplying matrices and vectors for the geometric transformations, but doing it on audio data is not very different, well, the difference is just that it's a lot lighter computational load, hence the small amount of GPU power it needs.
