Skip to content

Pull requests: NVIDIA/Fuser

New pull request New
134 Open 2,465 Closed
134 Open 2,465 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Fix communication lowering to support DID loop parallelization.
#3284 opened Oct 25, 2024 by wujingyue • Draft
Change shape of HSH_NT_128BSwizzle
#3281 opened Oct 25, 2024 by zasdfgbnm Loading…
1
Export DEBUG_SERDE=true in compare_codegen.sh
#3280 opened Oct 25, 2024 by jacobhinkle • Draft
1
[WIP] Schedule Hopper mma instruction matmul
#3278 opened Oct 25, 2024 by jacobhinkle • Draft
Refactor MultiMatmulSchedulers matmul Top-Down Matmul Dev
#3277 opened Oct 25, 2024 by rdspring1 Loading…
1
1
add knobs control inner dim unroll and outer dim unroll in pointwise scheduler
#3275 opened Oct 25, 2024 by liqiangxl • Draft
3
avoid treating pointer to bool as bool when handling kir::asm in codegen
#3274 opened Oct 25, 2024 by liqiangxl Loading…
4
check vectorization factor of shared memory consumers to avoid illegal vectorization size
#3271 opened Oct 25, 2024 by liqiangxl Loading…
3
Add enable_options and disable_options to fd.execute
#3270 opened Oct 24, 2024 by Priya2698 Loading…
2
expose and test getPersistentBufferStorageParams
#3264 opened Oct 24, 2024 by liqiangxl • Draft
4
Create dispatch system for executors
#3263 opened Oct 24, 2024 by csarofeen • Draft
24
patch potential segfault
#3262 opened Oct 24, 2024 by jjsjann123 Loading…
29
Lowering vectorized pad
#3261 opened Oct 23, 2024 by jjsjann123 • Draft
5
Sequence Parallel MLP
#3259 opened Oct 23, 2024 by cowanmeg Loading…
10
Pointwise shouldn't check transpose scheduler
#3256 opened Oct 22, 2024 by naoyam • Draft
8
Gather data for pointwise fusions Autotune Generate heuristics through machine learning models.
#3255 opened Oct 22, 2024 by rdspring1 • Draft
2
Introduce @retry_on_oom_or_skip_test
#3252 opened Oct 22, 2024 by jacobhinkle Loading…
11
[WIP] Cat scheduling
#3250 opened Oct 22, 2024 by naoyam • Draft
enable unroll on top of vectorization in pointwise scheduler
#3249 opened Oct 21, 2024 by liqiangxl • Draft
2
Host benchmarking for many-segment fusion
#3230 opened Oct 19, 2024 by Priya2698 Loading…
5
Adaptive layernorm host benchmark
#3229 opened Oct 19, 2024 by Priya2698 Loading…
4
tune heuristics for inner outer scheduler after cached inputs are moved to smem
#3223 opened Oct 18, 2024 by liqiangxl • Draft
3
always prioritize using smem to store inner persistent buffers
#3217 opened Oct 18, 2024 by liqiangxl Loading…
21
Stub out support for Hopper in matmul scheduler
#3216 opened Oct 18, 2024 by jacobhinkle • Draft
Use a single mbarrier to track all loads
#3202 opened Oct 16, 2024 by zasdfgbnm • Draft
1
ProTip! Filter pull requests by the default branch with base:main.