[CPU] Migrate unquantized MoE to the modular-kernel experts structure (#50133)

Signed-off-by: jiang1.li <jiang1.li@intel.com>
Signed-off-by: Fadi Arafeh <fadi.arafeh@arm.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
Co-authored-by: Fadi Arafeh <fadi.arafeh@arm.com>
This commit is contained in:
Li, Jiang
2026-08-03 14:55:04 +08:00
committed by GitHub
co-authored by Claude Opus 5 Fadi Arafeh
parent b3f97dae24
commit 0a6446005d
20 changed files with 972 additions and 839 deletions
+2 -1
View File
@@ -461,6 +461,7 @@ set(VLLM_EXT_SRC
"csrc/cpu/pos_encoding.cpp"
"csrc/cpu/mamba_cpu.cpp"
"csrc/moe/dynamic_4bit_int_moe_cpu.cpp"
"csrc/cpu/cpu_fused_moe.cpp"
"csrc/cpu/cpu_attn.cpp"
"csrc/cpu/torch_bindings.cpp")
@@ -478,7 +479,6 @@ if (ASIMD_FOUND AND NOT APPLE_SILICON_FOUND)
"csrc/cpu/cpu_tanhf_neon.hpp"
${VLLM_EXT_SRC})
if (ARM_BF16_FOUND)
set(VLLM_EXT_SRC "csrc/cpu/cpu_fused_moe.cpp" ${VLLM_EXT_SRC})
if (ARM_I8MM_FOUND)
set(VLLM_EXT_SRC "csrc/cpu/cpu_fused_moe_int8.cpp" ${VLLM_EXT_SRC})
endif()
@@ -535,6 +535,7 @@ if (ENABLE_X86_ISA)
set(VLLM_EXT_SRC_AVX2
"csrc/cpu/sgl-kernels/fla.cpp"
"csrc/cpu/cpu_fused_moe.cpp"
"csrc/cpu/utils.cpp"
"csrc/cpu/spec_decode_utils.cpp"
"csrc/cpu/cpu_attn.cpp"