David Huang 84778e9770 CUDA/HIP: Share the same unified memory allocation logic. (#12934) ... Replace compile-time `GGML_HIP_UMA` with environment variable `GGML_CUDA_ENABLE_UNIFIED_MEMORY`. This unifies the usage on NVIDIA and AMD GPUs, and allows a single binary to be shared between integrated and dedicated GPUs.		2025-04-15 11:20:38 +02:00
..
cuda.h	CUDA: add BF16 support (#11093)	2025-01-06 02:33:52 +01:00
hip.h	CUDA/HIP: Share the same unified memory allocation logic. (#12934)	2025-04-15 11:20:38 +02:00
musa.h	cuda : fix HIP and MUSA BF16 (#0)	2025-04-07 18:44:17 +03:00