Skip to content

Pull requests: ROCm/composable_kernel

New pull request New
82 Open 3,228 Closed
82 Open 3,228 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
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Add unit tests for template optimization helpers
#3610 opened Jan 19, 2026 by tenpercent Loading…
[CK TILE] Refactor sequence_reverse_inclusive_scan
#3609 opened Jan 19, 2026 by CongMa13 Loading…
7 tasks
Add build time optimization documentation
#3608 opened Jan 19, 2026 by tenpercent Loading…
19
Dlejeune/ck tile multi reduce improvements
#3607 opened Jan 19, 2026 by damien-lejeune Loading…
2 of 7 tasks
4
[CK_TILE] Add irregular tail vectorloads pipeline
#3606 opened Jan 19, 2026 by jakpiase Loading…
Re enable f8 x bf8 tests on compv3 and compv4
#3605 opened Jan 19, 2026 by SamiAario-AMD Loading…
3 of 7 tasks
[CK_BUILDER] Replace reference conv with old ck implementation
#3604 opened Jan 19, 2026 by Snektron Loading…
1
Support for a4w4 (fp4) in block scale gemm AB quant organization: streamhpc
#3603 opened Jan 19, 2026 by ErwinTerpstra • Draft
7 tasks
Mx fp6 flatmm
#3601 opened Jan 19, 2026 by eeezio Loading…
4 of 7 tasks
5
Replace nested static_for lambdas with compile-time search helper
#3600 opened Jan 16, 2026 by tenpercent Loading…
1 task
[CK TILE] remove dependency on std chrono
#3599 opened Jan 16, 2026 by CongMa13 Loading…
7 tasks
8
[DRAFT] Add support to fp16 + compute fp16 and bf16 + compute bf16 contractions
#3598 opened Jan 16, 2026 by evedovelli • Draft
7 tasks done
Rewrite O(N) recursive templates with O(1) pack expansion
#3596 opened Jan 16, 2026 by tenpercent Loading…
1 task
WMMA grouped conv fwd large tensor bias bnorm clamp organization: streamhpc
#3595 opened Jan 16, 2026 by wj-laskowski Loading…
6 of 7 tasks
1
[gfx906] Collection of fixes for MI50/MI60 (non-MFMA) GPUs
#3593 opened Jan 16, 2026 by dbsanfte • Draft
3
[CK_BUILDER] Add reflection for wmma and bwd weight instances to ck builder reflection
#3592 opened Jan 16, 2026 by kabrahamAMD • Draft
6
Add container and tuple optimization helpers
#3590 opened Jan 16, 2026 by tenpercent Loading…
1 task
Replace lambdas with named functors in transform_tensor_descriptor
#3589 opened Jan 16, 2026 by tenpercent Loading…
1 task
Add generate_identity_sequences helper for common pattern
#3588 opened Jan 16, 2026 by tenpercent Loading…
1 task
Rewrite StaticallyIndexedArray to use C-array instead of Tuple
#3587 opened Jan 16, 2026 by tenpercent • Draft
2 tasks done
3
[CK_TILE][FMHA] Add new tile size for async
#3586 opened Jan 16, 2026 by LJ-underdog Loading…
7 tasks done
3
Optimize sequence_gen and uniform_sequence_gen to reduce template instantiation depth
#3585 opened Jan 16, 2026 by tenpercent Loading…
1 task
4
WMMA grouped conv fwd large tensor extra flavors organization: streamhpc
#3582 opened Jan 15, 2026 by wj-laskowski Loading…
6 of 7 tasks
Remove code duplications in batched gemm wmma
#3580 opened Jan 15, 2026 by chris-tsiaousis-hpc • Draft
4 of 7 tasks
Fix HIP target attributes on deduction guides as they're deprecated
#3579 opened Jan 15, 2026 by aosewski Loading…
7 tasks
@aosewski
ProTip! no:milestone will show everything without a milestone.