Re:Generic GPU Kernels
by futharkshill on Tuesday December 07, 2021 @04:04AM
(#29470416)
I think it will always be slower than hand-optimized GPU code, just like assembly. But for most complex programs, I think the compiler is better than humans.
@arthas, the author, at some point made comparisons against implementations and it was a most twice as slow, but often faster.
How fast can Futhark be compared to a standard CUDA loop with a few arithmetic, load and save operations? Basically, suppose you're doing simple gathering and scattering?