|
ABACUS develop
Atomic-orbital Based Ab-initio Computation at UStc
|
Unified CUDA/ROCm API macros for portable GPU code. More...
Go to the source code of this file.
Unified CUDA/ROCm API macros for portable GPU code.
This header provides macro abstraction for CUDA/ROCm APIs, allowing a single implementation to work with both CUDA and ROCm backends.
Usage: #include "gpu_runtime.h" gpuError_t err = gpuGetDeviceCount(&count); if (err != gpuSuccess) { ... }