Skip to content

Pull requests: ROCm/aiter

New pull request New
199 Open 1,682 Closed
199 Open 1,682 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 ENABLE_CK=0 build option for Triton-only builds
#2062 opened Feb 19, 2026 by sunway513 Loading…
4 of 5 tasks
2
GFX1250 Kernels - GEMMa8w8 blockscale
#2060 opened Feb 18, 2026 by amirumoAMD • Draft
1 task
fix aiter.ops.triton.rope import
#2058 opened Feb 17, 2026 by Rohan138 Loading…
1 task
3
Enabling FPMX4 GEMM on non-FPMX4 devices (Navi31 in particular)
#2056 opened Feb 16, 2026 by ekuznetsov139 Loading…
3
Silence certain warnings stemming from CK
#2055 opened Feb 16, 2026 by Micky774 Loading…
1 task
2
Support per_block for Pa PS
#2053 opened Feb 14, 2026 by ZhangLirong-amd Loading…
1 task
10
Prepare repository for size optimization
#2052 opened Feb 14, 2026 by sunway513 Loading…
2
Add Model Benchmarking Tool
#2050 opened Feb 13, 2026 by lucas-santos-amd Loading…
1 task done
@lucas-santos-amd
[TRITON] Add smoothquant int8 MoE kernel
#2049 opened Feb 12, 2026 by nsusanto Loading…
10
GFX1250 Gluon MoE A4W4 Kernel
#2047 opened Feb 12, 2026 by farlukas • Draft
1 task
3
[TEST] Flash Attention Integration CI from fork ok-to-test
#2044 opened Feb 12, 2026 by micmelesse Loading…
1
upload mla_a8w8_qh64_qseqlen4_gqaratio16 co in MI300
#2042 opened Feb 12, 2026 by minmengdie Loading…
1 task
1
Introduce HipKittens based nhead=128 MLA Kernel
#2039 opened Feb 12, 2026 by ruanjm Loading…
2
gfx1250 gluon initial gemm for a8w8 MoE blockscale kernel
#2038 opened Feb 11, 2026 by nsusanto Loading…
1 task
gfx12 gemm a8w8
#2036 opened Feb 11, 2026 by ahmed-bsod • Draft
10
Top-K Top-P Sampling Kernel Optimization
#2034 opened Feb 11, 2026 by aryaman-gupta Loading…
[FIX] fix a16 causal mha bwd case for python api
#2029 opened Feb 11, 2026 by JaxChen29 Loading…
1 task
[MOE]: add qwen3-VL UT
#2028 opened Feb 11, 2026 by xudoyuan Loading…
1 task
[not ready] Fuse rms rope blk quant kernel
#2027 opened Feb 11, 2026 by yzhou103 Loading…
1 task
3
update ut
#2025 opened Feb 11, 2026 by amd-ruitang3 Loading…
1 task
9
Fix wrong path to the tune script
#2023 opened Feb 10, 2026 by amd-yashagar Loading…
5
feat(mla_prl_ps): optimize get_ps_metadata & enhance ut
#2019 opened Feb 10, 2026 by dbyoung18 Loading…
feat(ck_tile): add a8w8 blockscale gemm with preshuffleQuant support
#2018 opened Feb 10, 2026 by amd-khushbu • Draft
1 task
[TRITON] Add MXFP4 quantization support to triton unified attention kernel
#2012 opened Feb 9, 2026 by amd-xiaoyu12 • Draft
1 task
Update FMHA calls with new MX qscale related args
#2008 opened Feb 9, 2026 by ex-rzr • Draft
1 task done
ProTip! Adding no:label will show everything without a label.