Skip to content

[None][feat] add CuteDSL FP8/FP16 MLA decode attention fmha lib #15138

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
haow-nv wants to merge 21 commits into
base: main
Choose a base branch
from
Open

[None][feat] add CuteDSL FP8/FP16 MLA decode attention fmha lib #15138

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
21 commits
Select commit Hold shift + click to select a range
e2d231b
[None][feat] add CuteDSL FP8/FP16 MLA decode attention backend
haow-nv Jun 8, 2026
e6126e4
[None][feat] CuteDSL MLA decode: fp8 scaling fix, BF16 support, per-l…
haow-nv Jun 12, 2026
bdb173b
[None][fix] CuteDSL MLA decode: v1 paged-KV layout + CUDA-graph-safe …
haow-nv Jun 16, 2026
8c90cf6
[TRTLLM-12807][feat] Wrap CuteDSL MLA decode as FMHA lib
yuxianq Jun 13, 2026
020cbd8
[None][feat] CuteDSL MLA decode: MTP causal mask + can_implement gate
haow-nv Jun 24, 2026
6748399
[None][feat] CuteDSL MLA decode: persistent + split-KV alignment (fla…
haow-nv Jun 29, 2026
f88b6ab
[None][feat] CuteDSL MLA decode: mixed prefill+decode page-table offs…
haow-nv Jun 30, 2026
03b9a27
[None][feat] CuteDSL MLA decode: autotuned is_persistent tactic, perf…
haow-nv Jul 6, 2026
b613a36
[None][chore] CuteDSL MLA: drop stale is_available comment
haow-nv Jul 6, 2026
d6d09f2
[None][chore] CuteDSL MLA decode: comment cleanup, drop default_is_pe…
haow-nv Jul 6, 2026
101d86d
[None][chore] CuteDSL MLA decode: apply pre-commit formatting, fix co…
haow-nv Jul 6, 2026
de0d156
[None][chore] CuteDSL MLA decode: drop standalone decode unit test
haow-nv Jul 7, 2026
a1a3cd8
[None][fix] CuteDSL MLA: reject sparse attention in FMHA gate
haow-nv Jul 7, 2026
fe54ced
[None][chore] CuteDSL MLA decode: trim comments, drop unused runner i…
haow-nv Jul 7, 2026
b0e0662
[None][perf] CuteDSL MLA: gate (128,1) decode on spec-decode off
haow-nv Jul 7, 2026
dcd5b37
[None][perf] CuteDSL MLA: batch-aware perf gate, autotuner decode war…
haow-nv Jul 10, 2026
c6f842b
[None][chore] CuteDSL MLA: apply pre-commit formatting
haow-nv Jul 10, 2026
6cab399
[None][perf] CuteDSL MLA: restructure decode perf gate, drop (128,1) …
haow-nv Jul 14, 2026
f1476d7
[None][fix] stabilize CuteDSL MLA split-KV workspace for CUDA graphs
haow-nv Jul 15, 2026
0a97d45
[None][chore] CuteDSL MLA decode: drop debug kernel-arg dumps, fix au…
haow-nv Jul 23, 2026
f842131
Update registry.py
haow-nv Jul 23, 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
10 changes: 10 additions & 0 deletions .pre-commit-config.yaml
View file
Open in desktop

@yuxianq yuxianq Jul 24, 2026

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

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

For these 5 new files, we can apply new ruff fules for them. We can revert all changes in .pre-commit-config.yaml, legacy-files.txt, pyproject.toml and ruff-legacy.toml.

@haow-nv haow-nv Jul 24, 2026

Copy link
Copy Markdown
Collaborator Author

Choose a reason for hiding this comment

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

Thanks. Will revert all changes.

Original file line number Diff line number Diff line change
Expand Up @@ -271,6 +271,11 @@ common-files: &common_files |
tensorrt_llm/_torch/custom_ops/userbuffers_custom_ops.py |
tensorrt_llm/_torch/cute_dsl_kernels/__init__.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/__init__.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/__init__.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/__init__.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/mla_decode_fp16.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/mla_decode_fp8.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/mla_helpers.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/custom_pipeline.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/dense_blockscaled_gemm_persistent.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/utils.py |
Expand Down Expand Up @@ -1578,6 +1583,11 @@ legacy-files: &legacy_files |
tensorrt_llm/_torch/custom_ops/userbuffers_custom_ops.py |
tensorrt_llm/_torch/cute_dsl_kernels/__init__.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/__init__.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/__init__.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/__init__.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/mla_decode_fp16.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/mla_decode_fp8.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/mla_helpers.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/custom_pipeline.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/dense_blockscaled_gemm_persistent.py |
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/utils.py |
Expand Down
5 changes: 5 additions & 0 deletions legacy-files.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -263,6 +263,11 @@ tensorrt_llm/_torch/custom_ops/trtllm_gen_custom_ops.py
tensorrt_llm/_torch/custom_ops/userbuffers_custom_ops.py
tensorrt_llm/_torch/cute_dsl_kernels/__init__.py
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/__init__.py
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/__init__.py
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/__init__.py
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/mla_decode_fp16.py
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/mla_decode_fp8.py
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/mla_helpers.py
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/custom_pipeline.py
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/dense_blockscaled_gemm_persistent.py
tensorrt_llm/_torch/cute_dsl_kernels/blackwell/utils.py
Expand Down
5 changes: 5 additions & 0 deletions pyproject.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -320,6 +320,11 @@ exclude = [
"tensorrt_llm/_torch/custom_ops/userbuffers_custom_ops.py",
"tensorrt_llm/_torch/cute_dsl_kernels/__init__.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/__init__.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/__init__.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/__init__.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/mla_decode_fp16.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/mla_decode_fp8.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/mla_helpers.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/custom_pipeline.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/dense_blockscaled_gemm_persistent.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/utils.py",
Expand Down
5 changes: 5 additions & 0 deletions ruff-legacy.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -280,6 +280,11 @@ include = [
"tensorrt_llm/_torch/custom_ops/userbuffers_custom_ops.py",
"tensorrt_llm/_torch/cute_dsl_kernels/__init__.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/__init__.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/__init__.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/__init__.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/mla_decode_fp16.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/mla_decode_fp8.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/attention/mla/mla_helpers.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/custom_pipeline.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/dense_blockscaled_gemm_persistent.py",
"tensorrt_llm/_torch/cute_dsl_kernels/blackwell/utils.py",
Expand Down
2 changes: 2 additions & 0 deletions tensorrt_llm/_torch/attention_backend/fmha/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,7 @@
# See the License for the specific language governing permissions and
# limitations under the License.

from .cute_dsl import CuteDslMlaFmha
from .fallback import FallbackFmha
from .flashinfer_trtllm_gen import FlashInferTrtllmGenFmha
from .interface import Fmha
Expand All @@ -23,6 +24,7 @@
__all__ = [
"DEFAULT_FMHA_LIBS",
"FMHA_LIBS",
"CuteDslMlaFmha",
"FallbackFmha",
"FlashInferTrtllmGenFmha",
"Fmha",
Expand Down
Loading
Loading