perf: optimize fused FP8 GEMM kernel

2026-08-18 19:56:15 +08:00
parent 1bcd8f53ab
commit cb51a3587b
5 changed files with 769 additions and 380 deletions
@@ -61,9 +61,6 @@ foreach(name ${KERNELS})
        "${PYTHON_INCLUDE_DIR}")

    target_link_libraries(${name} PRIVATE ${TORCH_LIBS})
-    if(${name} STREQUAL "fp8_mm")
-        target_link_libraries(${name} PRIVATE CUDA::cublasLt)
-    endif()
    target_link_options(${name} PRIVATE "-Wl,-rpath,${TORCH_LIB_DIR}")

    target_compile_options(${name} PRIVATE