Not directly. AMD GPUs do not execute CUDA programs as they are. AMD's route is source translation: HIPIFY rewrites CUDA code into HIP code, which then compiles for AMD hardware.
What HIPIFY does
According to AMD's HIPIFY documentation, it ships two tools. hipify-clang parses CUDA code and converts it to HIP code. hipify-perl swaps CUDA API calls for HIP equivalents. They handle many runtime API calls, kernel launch syntax and keywords such as __global__ and __device__.
The limits are stated plainly: "HIPIFY does not automatically convert all CUDA code into HIP code seamlessly." CUDA libraries and third-party libraries with no HIP equivalent cannot be translated, and converted code still needs review, testing and tuning on AMD GPUs. The documentation describes translating source, not running compiled CUDA binaries.
What AMD says about drop-in use
AMD's HIP documentation says "HIP is not intended to be a drop-in replacement for NVIDIA CUDA." It adds that libraries such as hipFFT and hipBLAS "provide API compatibility with equivalent NVIDIA CUDA libraries", which eases a port.
Next, read what HIP is, then ROCm vs CUDA. The broader comparison is in NVIDIA's main competitors in AI.
