Skip to content

feat: add SM120 fmha_v2 kernels to AOT pip wheel builds #2885

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
blake-snc wants to merge 2 commits into
base: main
Choose a base branch
from
Open

feat: add SM120 fmha_v2 kernels to AOT pip wheel builds #2885

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
429d8d2
feat: add SM120 fmha_v2 kernels to AOT pip wheel builds
blake-snc Mar 24, 2026
0285211
nit: update SM120 comment to include attention kernels
blake-snc Mar 24, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 3 additions & 1 deletion flashinfer/aot.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -44,6 +44,7 @@
gen_single_decode_module,
gen_single_prefill_module,
gen_trtllm_gen_fmha_module,
gen_trtllm_fmha_v2_sm120_module,
)
from .jit.cascade import gen_cascade_module
from .jit.cpp_ext import get_cuda_version
Expand Down Expand Up @@ -522,13 +523,14 @@ def gen_all_modules(
if has_sm121:
jit_specs.append(gen_fp4_quantization_sm121_module())
if has_sm120 or has_sm121:
# SM120 and SM121 share the same CUTLASS kernels for fused MOE and GEMM.
# SM120 and SM121 share the same kernels for fused MOE, GEMM, and attention.
# The SM120 module generators use supported_major_versions=[12] which
# compiles for all SM12x targets.
jit_specs.append(gen_cutlass_fused_moe_sm120_module())
jit_specs.append(gen_gemm_sm120_module())
jit_specs.append(gen_gemm_sm120_module_cutlass_fp4())
jit_specs.append(gen_gemm_sm120_module_cutlass_mxfp8())
jit_specs.append(gen_trtllm_fmha_v2_sm120_module())
Copy link
Copy Markdown
Contributor

@gemini-code-assist gemini-code-assist bot Mar 24, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

medium

With the addition of this fmha_v2 module, the comment on lines 525-527 is now slightly outdated as it only mentions 'fused MOE and GEMM'. For better maintainability, please consider updating it to include attention kernels for clarity.

For example:

-            # SM120 and SM121 share the same CUTLASS kernels for fused MOE and GEMM.
+            # SM120 and SM121 share the same kernels for fused MOE, GEMM, and attention.

if has_sm120f:
jit_specs.append(gen_fp4_quantization_sm120f_module())

Expand Down
Loading