Lumpiasty/llama.cpp
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

HIP: RDNA4 tensor core support for MMF (#17077)

Browse Source 
* mmf for rdna4

* align the padding for rdna4

* forbit mul_mat_f for rdna4

* fix as comment

* remove device kernels

* add constexpr for early return

* update based on review comment

* change based on the review comment

* pass compile error

* keep code consistency

---------

Co-authored-by: zhang hui <you@example.com>
This commit is contained in:
yulo
2025-11-22 00:03:24 +01:00
committed by GitHub
co-authored by zhang hui
parent 8e9ddba610
commit 028f93ef98
5 changed files with 180 additions and 23 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-cuda/convert.cuh
+9
View File
@@ -39,6 +39,15 @@ template<typename dst_t, typename src_t>
return __float2bfloat16(float(x));
} else if constexpr(std::is_same_v<src_t, nv_bfloat16>) {
return __bfloat162float(x);
} else if constexpr(std::is_same_v<src_t, float2> && std::is_same_v<dst_t, half2>) {
return __float22half2_rn(x);
} else if constexpr(std::is_same_v<src_t, float2> && std::is_same_v<dst_t, nv_bfloat162>) {
// bypass compile error on cuda 12.0.1
#ifdef GGML_USE_HIP
return __float22bfloat162_rn(x);
#else
return {x.x, x.y};
#endif // GGML_USE_HIP
} else if constexpr(std::is_same_v<dst_t, int32_t>) {
return int32_t(x);
} else {