On 5/7/19 1:59 PM, Prentice Bisbal via Beowulf wrote:
I agree. That means a LOT of codes will have to be ported from CUDA to whatever AMD uses. I know AMD announced their HIP interface to convert CUDA code into something that will run on AMD processors, but I don't know how well that works in theory. Frankly, I haven't heard anything about it since it was announced at SC a few years ago.
You can always try it yourself - a fun experiment might be to see if you can use hipify to convert a CUDA app to HIP and then build it (and see how much extra work is needed to finish the port if it doesn't get it all) and then benchmark the original against the HIP version on an nVidia GPU to confirm it still works correctly and if there's a performance penalty/gain.
https://github.com/ROCm-Developer-Tools/HIP cheers! Chris -- Chris Samuel : http://www.csamuel.org/ : Berkeley, CA, USA _______________________________________________ Beowulf mailing list, Beowulf@beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit https://beowulf.org/cgi-bin/mailman/listinfo/beowulf
