Re:Ask HN: Can competitors catch up to Apple Silicon?
by Const-me on Tuesday July 05, 2022 @02:20PM
(#31991824)
DirectCompute is pretty close to CUDA. When handled with care, it can do these compute-heavy pieces with similar performance.
The main downside is the libraries: cuBLAS, cuFFT, and similar pieces of CUDA are manually optimized, and nVidia ships multiple versions of these compute kernels for different GPUs. DirectCompute doesn’t have an equivalent, need to implement manually in HLSL. This can be tricky because for optimal performance different chips need slightly different implementations.
The main downside is the libraries: cuBLAS, cuFFT, and similar pieces of CUDA are manually optimized, and nVidia ships multiple versions of these compute kernels for different GPUs. DirectCompute doesn’t have an equivalent, need to implement manually in HLSL. This can be tricky because for optimal performance different chips need slightly different implementations.