I need to do a lot of bit-wise operations on GPUs, but cannot find any information regarding whether NVIDIA GPU hardware is big or little-endian.
All of the supported CUDA platforms use little-endian CPUs, and cudaMemcpy() can copy data structures to the device without knowing the data format, so I would assume the GPU is also little-endian. The GPU might support both big and little endian execution (as some CPUs also do this) as a hedge against future CUDA platforms being big endian.
My guess is the answer has to be either "little-endian" or "both".
NVIDIA GPUs are little-endian, per the Hardware Implementation section of the CUDA guide.
All of the supported CUDA platforms use little-endian CPUs, and cudaMemcpy() can copy data structures to the device without knowing the data format, so I would assume the GPU is also little-endian. The GPU might support both big and little endian execution (as some CPUs also do this) as a hedge against future CUDA platforms being big endian.
My guess is the answer has to be either "little-endian" or "both".
© 2022 - 2024 — McMap. All rights reserved.