public_git
/
whisper.cpp
mirror of https://github.com/ggml-org/whisper.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

sycl : support to malloc memory on device more than 4GB, update the doc and script (llama/17566) 

Co-authored-by: Neo Zhang Jianyu <jianyu.zhang@intel.com>
This commit is contained in:
Neo Zhang 2025-11-29 20:59:44 +08:00 committed by Georgi Gerganov
parent 28dff06555
commit a3459484bf
No known key found for this signature in database
GPG Key ID: 449E073F9DC10735
2 changed files with 6 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
ggml/src/ggml-sycl/CMakeLists.txt
View File

@ -91,7 +91,10 @@ if (GGML_SYCL_F16)
add_compile_definitions(GGML_SYCL_F16)
endif()
if (GGML_SYCL_TARGET STREQUAL "NVIDIA")
if (GGML_SYCL_TARGET STREQUAL "INTEL")
add_compile_definitions(GGML_SYCL_WARP_SIZE=16)
target_link_options(ggml-sycl PRIVATE -Xs -ze-intel-greater-than-4GB-buffer-required)
elseif (GGML_SYCL_TARGET STREQUAL "NVIDIA")
add_compile_definitions(GGML_SYCL_WARP_SIZE=32)
elseif (GGML_SYCL_TARGET STREQUAL "AMD")
# INFO: Allowed Sub_group_sizes are not consistent through all
@ -100,7 +103,8 @@ elseif (GGML_SYCL_TARGET STREQUAL "AMD")
# Target archs tested working: gfx1030, gfx1031, (Only tested sub_group_size = 32)
add_compile_definitions(GGML_SYCL_WARP_SIZE=32)
else()
add_compile_definitions(GGML_SYCL_WARP_SIZE=16)
# default for other target
add_compile_definitions(GGML_SYCL_WARP_SIZE=32)
endif()
if (GGML_SYCL_GRAPH)

3
ggml/src/ggml-sycl/cpy.cpp
View File

@ -515,9 +515,6 @@ void ggml_sycl_cpy(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, co
const int64_t ne = ggml_nelements(src0);
GGML_ASSERT(ne == ggml_nelements(src1));
GGML_ASSERT(ggml_nbytes(src0) <= INT_MAX);
GGML_ASSERT(ggml_nbytes(src1) <= INT_MAX);
GGML_TENSOR_BINARY_OP_LOCALS01;
SYCL_CHECK(ggml_sycl_set_device(ctx.device));