Re:Lisa Su Saved AMD. Now She Wants Nvidia's AI Crown
by KeplerBoy on Friday June 02, 2023 @04:29PM
(#36170338)
True.
And the resulting PTX is worse when it's generated from OpenCL C instead of CUDA C. I tested that recently with a toy FFT kernel and the CUDA pipeline produced a lot more efficient FMA instructions.
And the resulting PTX is worse when it's generated from OpenCL C instead of CUDA C. I tested that recently with a toy FFT kernel and the CUDA pipeline produced a lot more efficient FMA instructions.