Phil J Taylor
Joined: 26 Dec 09
Has anyone considered adding support for the various GPU's to FFTw?
What I am 'imagining' might make it easier to add support for ATI and various other GPU's.
Doing it this way would never 'out perform' the optimized CUDA apps that we have since they 'reserve' the whole GPU but, it seems like it might significantly boost 2, 4,or 8 non-CUDA application streams. Also, it might increase the overall utilization of the available GPU's and, perhaps, allow our apps to benefit from GPU's with smaller memories.
Any thoughts? Has it been tried? Any pointers to other discussions about anything like this?
©2016 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.