biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Kernel] Add NVFP4 MoE CUTLASS support for SM120 (#29242)

Browse Source 
Signed-off-by: mgoin <mgoin64@gmail.com>
Signed-off-by: Michael Goin <mgoin64@gmail.com>
This commit is contained in:
Michael Goin
2025-11-25 09:59:07 -05:00
committed by GitHub
parent dbc3d9991a
commit e502098643
8 changed files with 264 additions and 30 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
csrc/quantization/fp4/nvfp4_quant_entry.cu
View File

@@ -24,8 +24,9 @@ void scaled_fp4_quant_sm1xxa(torch::Tensor const& output,
torch::Tensor const& input_sf);
#endif
#if defined ENABLE_NVFP4_SM100 && ENABLE_NVFP4_SM100
void scaled_fp4_experts_quant_sm100a(
#if (defined(ENABLE_NVFP4_SM100) && ENABLE_NVFP4_SM100) || \
(defined(ENABLE_NVFP4_SM120) && ENABLE_NVFP4_SM120)
void scaled_fp4_experts_quant_sm1xxa(
torch::Tensor& output, torch::Tensor& output_scale,
torch::Tensor const& input, torch::Tensor const& input_global_scale,
torch::Tensor const& input_offset_by_experts,
@@ -54,8 +55,9 @@ void scaled_fp4_experts_quant(
torch::Tensor const& input, torch::Tensor const& input_global_scale,
torch::Tensor const& input_offset_by_experts,
torch::Tensor const& output_scale_offset_by_experts) {
#if defined ENABLE_NVFP4_SM100 && ENABLE_NVFP4_SM100
return scaled_fp4_experts_quant_sm100a(
#if (defined(ENABLE_NVFP4_SM100) && ENABLE_NVFP4_SM100) || \
(defined(ENABLE_NVFP4_SM120) && ENABLE_NVFP4_SM120)
return scaled_fp4_experts_quant_sm1xxa(
output, output_scale, input, input_global_scale, input_offset_by_experts,
output_scale_offset_by_experts);
#endif