vLLM/csrc/cpu at 4e4eab73ed2b4e00931a4636e926b8df1cb1fe50 - vLLM

micro_gemm

[HARDWARE][POWER] Enable W8A8 INT8 MoE on POWER (#55316 )

2026-09-17 21:45:13 +08:00

sgl-kernels

[Bugfix][CPU] Fix macOS build on Apple Clang < 17: structured binding… (#58932 )

2026-10-02 10:01:07 +08:00

activation_lut_bf16.cpp

[perf][cpu] Accelerate BF16 GELU with LUT impl on Arm CPUs (#37469 )

2026-04-15 22:26:17 -07:00

activation.cpp

[CPU][Perf]Added tanh AOR for faster gelu activations. (#44639 )

2026-06-30 23:24:40 -07:00

cpu_arch_macros.h

[CPU][s390x] Optimize inference perf and add oneDNN INT8 GEMM for s390x (#50219 )

2026-07-31 17:49:31 +08:00

cpu_attn_amx_fp8.hpp

[CPU] [Feat] Add native AMX-FP8 attention impl for Diamond Rapids (#49410 )

2026-09-06 11:06:56 +08:00

cpu_attn_amx.hpp

[Feat] CPU fp8 attn for AMX/AVX-512 (#39445 )

2026-04-29 20:43:21 +08:00

cpu_attn_fp8.hpp

[Refactor] Remove kernel dead code (#51051 )

2026-08-07 01:20:19 -07:00

cpu_attn_impl.hpp

[Misc] Consolidate csrc shared headers and prune dead code (#60721 )

2026-10-12 09:46:22 -07:00

cpu_attn_neon_bfmmla.hpp

[CPU][Perf] refactor paged attention for Arm CPUs (#56045 )

2026-09-20 12:25:13 +08:00

cpu_attn_neon.hpp

[CPU][Perf] refactor paged attention for Arm CPUs (#56045 )

2026-09-20 12:25:13 +08:00

cpu_attn_rvv.hpp

[CPU][RISC-V] Add VLEN=256 support to RVV attention kernels (#42943 )

2026-05-21 04:50:49 -07:00

cpu_attn_vec16.hpp

[Feat] CPU fp8 attn for AMX/AVX-512 (#39445 )

2026-04-29 20:43:21 +08:00

cpu_attn_vec.hpp

[Bugfix][CPU] Fix Clang OpenMP build on macOS (#49021 )

2026-07-21 09:58:52 +00:00

cpu_attn_vsx.hpp

[HARDWARE][POWER] optimize math functions of VSX power (#47321 )

2026-07-07 09:35:47 +00:00

cpu_attn_vxe.hpp

[Feat] CPU fp8 attn for AMX/AVX-512 (#39445 )

2026-04-29 20:43:21 +08:00

cpu_attn.cpp

[Bugfix][DiffusionGemma] Read the causal mask as int32 and compile the sample step (#59992 )

2026-10-06 19:25:51 +08:00

cpu_fused_moe_activations.hpp

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

2026-08-03 14:55:04 +08:00

cpu_fused_moe_int8.cpp

[HARDWARE][POWER] Enable W8A8 INT8 MoE on POWER (#55316 )

2026-09-17 21:45:13 +08:00

cpu_fused_moe.cpp

[Bugfix][CPU] Fix several bugs (#54042 )

2026-08-31 15:33:06 +08:00

cpu_isa.cpp

[CPU] Add CPU FP8 W8A8 linear/MoE support (#49942 )

2026-09-18 17:59:42 +08:00

cpu_tanhf_neon.hpp

[CPU][Perf]Added tanh AOR for faster gelu activations. (#44639 )

2026-06-30 23:24:40 -07:00

cpu_types_arm.hpp

[CPU] Add MLA backend so DeepSeek-V2/V3 can run on CPU (#49453 )

2026-08-06 16:51:41 +08:00

cpu_types_riscv_defs.hpp

[CPU][RISC-V] Add RVV path for W4A8 INT4 GEMM (#45269 )

2026-06-25 08:18:10 +00:00

cpu_types_riscv_impl.hpp

[Bugfix][CPU][RISC-V] Fix build: make FP32Vec copy constructors non-explicit (#51099 )

2026-08-14 04:59:01 +00:00

cpu_types_riscv.hpp

[CPU][RISC-V] Support multiple RVV VLEN targets via compile-time dispatch (#39478 )

2026-04-20 14:36:59 +08:00

cpu_types_scalar.hpp

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

2026-08-03 14:55:04 +08:00

cpu_types_vsx.hpp

[Hardware][Power] Enable W4A16 (AWQ & GPTQ) quantization on POWER10 using VSX (#59149 )

2026-09-30 19:44:48 +08:00

cpu_types_vxe.hpp

[CPU] Enable GPTQ and AWQ quantization for s390x (#51148 )

2026-08-10 11:41:43 +08:00

cpu_types_x86.hpp

[CPU] [Feat] Add native AMX-FP8 attention impl for Diamond Rapids (#49410 )

2026-09-06 11:06:56 +08:00

cpu_types.hpp

[HARDWARE][POWER] optimize math functions of VSX power (#47321 )

2026-07-07 09:35:47 +00:00

cpu_wna16.cpp

[Hardware][Power] Enable W4A16 (AWQ & GPTQ) quantization on POWER10 using VSX (#59149 )

2026-09-30 19:44:48 +08:00

dnnl_helper.cpp

[CPU] Speedup LM Head on Arm CPUs (#55352 )

2026-09-11 16:21:14 +08:00

dnnl_helper.h

[CPU] Speedup LM Head on Arm CPUs (#55352 )

2026-09-11 16:21:14 +08:00

dnnl_kernels.cpp

[CPU] Speedup LM Head on Arm CPUs (#55352 )

2026-09-11 16:21:14 +08:00

dynamic_4bit_int_moe_cpu.cpp

[Misc] Consolidate csrc shared headers and prune dead code (#60721 )

2026-10-12 09:46:22 -07:00

float_convert.hpp

Use std::bit_cast for type punning in CPU kernels (#45089 )

2026-06-10 22:07:44 -07:00

generate_cpu_attn_dispatch.py

Add pydocstyle to the ruff rules (#52136 )

2026-09-16 16:44:29 +00:00

layernorm.cpp

[Kernel] Add weightless RMSNorm CUDA kernels for has_weight=False (#41430 ) (#44109 )

2026-06-16 23:45:55 -07:00

mamba_cpu.cpp

[Hardware][CPU] Enable granite-4 model on cpu (#47641 )

2026-07-20 06:15:16 +00:00

mamba_kernels.hpp

[Hardware][CPU] Enable granite-4 model on cpu (#47641 )

2026-07-20 06:15:16 +00:00

mla_decode.cpp

[CPU] Add MLA backend so DeepSeek-V2/V3 can run on CPU (#49453 )

2026-08-06 16:51:41 +08:00

ops.h

[Misc] Consolidate csrc shared headers and prune dead code (#60721 )

2026-10-12 09:46:22 -07:00

pos_encoding.cpp

[HARDWARE][POWER] Enable fp16 support for PowerPC (#46135 )

2026-06-23 13:24:49 +00:00

sampling_kernels.cpp

[Bugfix][CPU] Fix fused Gumbel-max wrapping-window bias (#60221 )

2026-10-08 15:15:53 +08:00

shm.cpp

[HARDWARE][POWER] Enable SHM communicator support for PowerPC (#43754 )

2026-06-02 18:06:32 +08:00

spec_decode_utils.cpp

[CPU][Spec Decode] Enable DFlash SD for CPU (#44029 )

2026-06-25 15:32:48 +08:00

torch_bindings.cpp

[Misc] Consolidate csrc shared headers and prune dead code (#60721 )

2026-10-12 09:46:22 -07:00

utils.cpp

[CPU] Refactor CPU affinity and memory management (#39781 )

2026-04-17 21:01:08 +08:00

utils.hpp

[Hardware][Power] Unqualized MoE Backend for Power (VSX) (#51624 )

2026-08-13 10:56:09 +08:00