Refactor JIT compilation (+NVRTC support)#91
Closed
lucifer1004 wants to merge 14 commits into
Closed
Conversation
Signed-off-by: Zihua Wu <13583761+lucifer1004@users.noreply.github.com>
Signed-off-by: Zihua Wu <13583761+lucifer1004@users.noreply.github.com>
Signed-off-by: Zihua Wu <13583761+lucifer1004@users.noreply.github.com>
Collaborator
|
Thanks for this contribution, this is huge! Will take some time to review. |
Signed-off-by: Zihua Wu <13583761+lucifer1004@users.noreply.github.com>
Signed-off-by: Zihua Wu <13583761+lucifer1004@users.noreply.github.com>
Signed-off-by: Zihua Wu <13583761+lucifer1004@users.noreply.github.com>
Signed-off-by: Gabriel Wu <13583761+lucifer1004@users.noreply.github.com>
feat: add compat for older drivers and Windows
Signed-off-by: Zihua Wu <13583761+lucifer1004@users.noreply.github.com>
This comment was marked as outdated.
This comment was marked as outdated.
Signed-off-by: Zihua Wu <13583761+lucifer1004@users.noreply.github.com>
Signed-off-by: Zihua Wu <13583761+lucifer1004@users.noreply.github.com>
Collaborator
Author
|
It turned out that even older drivers can work with a newer CUDA container (my previous attempt was to use a newer CUDA installation from conda with an old driver), so I reverted to the original implementation of getting kernels. |
Contributor
|
This is an excellent work. May I ask if there are any plans regarding when it will be merged? @LyricZhao |
Collaborator
I plan to merge after the 5.1 holiday. |
Collaborator
Author
|
Moved to #94 |
LyricZhao
added a commit
that referenced
this pull request
Apr 16, 2026
* Working example * enhancement: apply advance tech * Fix: fix the swizzling limiting problem * Fix: fix the test problem * fix: diff the names of tma_copy * fix: fix test scripts * refa: clean up code * refa: use template tma copy * feat: run all tests * fix: cleanup code * fix: cleanup code * fix: add assert statement * fix: space elimination * Some code lints * Fix: remove one useless inst * Remove the parallel test script --------- Co-authored-by: Chenggang Zhao <chenggangz@deepseek.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
With this PR, the JIT compilation time is reduced by ~60% when using NVCC, ~80% when using NVRTC w/o PCH, and ~90% when using NVRTC w/ PCH.
Benchmark on Xeon(R) Platinum 8480C + H100 HBM:
Note that there is some perf drop when using NVRTC due to a known bug of NVRTC which leads to extra instructions (but in the
m=4096,n=2112,k=7168case, NVRTC version was faster, which was a bit strange). So NVCC is kept as the default compiler for now, while NVRTC can be enabled with extra env varDG_JIT_USE_NVRTC.