Skip to content

FA2 8.0 PTX#69

Merged
LucasWilkinson merged 1 commit intomainfrom
lwilkinson/fa2-ptx
Jun 16, 2025
Merged

FA2 8.0 PTX#69
LucasWilkinson merged 1 commit intomainfrom
lwilkinson/fa2-ptx

Conversation

@LucasWilkinson
Copy link
Collaborator

@LucasWilkinson LucasWilkinson commented Jun 9, 2025

NOTE: cmake changes lifted from vllm-project/vllm#18155

Helps with wheel size, not seeing any performance degradation on Blackwell once the cache is warmed up

See vllm-project/vllm#19336 for more details

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
@LucasWilkinson LucasWilkinson changed the title [WIP] FA2 8.0 PTX FA2 8.0 PTX Jun 16, 2025
@LucasWilkinson LucasWilkinson marked this pull request as ready for review June 16, 2025 16:09
@LucasWilkinson LucasWilkinson merged commit 763ad15 into main Jun 16, 2025
3 of 4 checks passed
LucasWilkinson added a commit that referenced this pull request Jun 16, 2025
Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
LucasWilkinson added a commit that referenced this pull request Jun 16, 2025
* varlen combine scheduler

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* cleanup

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* move check

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* standard scheduling algo

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* better heuristic

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* better comments

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* cleanup

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* cleanup

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* put in a more readable heurisitic

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* Apply suggestions from code review

Co-authored-by: Tyler Michael Smith <tysmith@redhat.com>
Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* FA2 8.0 PTX (#69)

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

---------

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
Co-authored-by: Tyler Michael Smith <tysmith@redhat.com>
zyongye pushed a commit to zyongye/flash-attention that referenced this pull request Aug 7, 2025
Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
zyongye pushed a commit to zyongye/flash-attention that referenced this pull request Aug 7, 2025
* varlen combine scheduler

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* cleanup

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* move check

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* standard scheduling algo

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* better heuristic

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* better comments

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* cleanup

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* cleanup

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* put in a more readable heurisitic

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* Apply suggestions from code review

Co-authored-by: Tyler Michael Smith <tysmith@redhat.com>
Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* FA2 8.0 PTX (vllm-project#69)

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

---------

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
Co-authored-by: Tyler Michael Smith <tysmith@redhat.com>
LucasWilkinson added a commit that referenced this pull request Aug 7, 2025
Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
LucasWilkinson added a commit that referenced this pull request Aug 7, 2025
* varlen combine scheduler

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* cleanup

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* move check

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* standard scheduling algo

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* better heuristic

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* better comments

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* cleanup

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* cleanup

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* put in a more readable heurisitic

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* Apply suggestions from code review

Co-authored-by: Tyler Michael Smith <tysmith@redhat.com>
Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* FA2 8.0 PTX (#69)

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

---------

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
Co-authored-by: Tyler Michael Smith <tysmith@redhat.com>
jayhshah pushed a commit that referenced this pull request Aug 8, 2025
Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
Signed-off-by: Jay Shah <jayhshah@gmail.com>
jayhshah pushed a commit that referenced this pull request Aug 8, 2025
* varlen combine scheduler

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* cleanup

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* move check

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* standard scheduling algo

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* better heuristic

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* better comments

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* cleanup

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* cleanup

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* put in a more readable heurisitic

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* Apply suggestions from code review

Co-authored-by: Tyler Michael Smith <tysmith@redhat.com>
Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

* FA2 8.0 PTX (#69)

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

---------

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
Co-authored-by: Tyler Michael Smith <tysmith@redhat.com>
Signed-off-by: Jay Shah <jayhshah@gmail.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants