Skip to content

Pull requests: ROCm/aiter

New pull request New
116 Open 1,013 Closed
116 Open 1,013 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

topk per row kernel
#1262 opened Oct 25, 2025 by ukannika Loading…
[TRITON] Instruction shape fix for Gluon gemm_a8w8_blockscale kernel
#1261 opened Oct 24, 2025 by eky-amd Loading…
Kernel naming: add reusable constexpr repr helper and testing it with gemm_a16w16
#1260 opened Oct 24, 2025 by Boss2002n • Draft
Fix rowwise a8w8 gemm in swizzled hipb_mm
#1258 opened Oct 24, 2025 by kliuae-amd Loading…
1 task
1
Fused GEMM + SILU for Llama4 Maverick
#1257 opened Oct 24, 2025 by juuso-oskari • Draft
CI: Add pre-check status check
#1252 opened Oct 23, 2025 by gyohuangxin • Draft
Fused TopK and Sigmoid kernel
#1251 opened Oct 23, 2025 by samremes • Draft
1 task
[WIP][TRITON] MHA PA optimizations
#1245 opened Oct 23, 2025 by cagrikymk • Draft
Revert "integrate cktile moe into fused_moe interface"
#1244 opened Oct 23, 2025 by dllehr-amd Loading…
add a tuned fp4 gemm ds config and insert entries in untuned config
#1243 opened Oct 22, 2025 by hongxiayang Loading…
1
fix the problem that v3's performance is worse than ck's
#1237 opened Oct 22, 2025 by minmengdie Loading…
1 task
5
Mla splitkv enhance split alg inte
#1233 opened Oct 21, 2025 by valarLip Loading…
1 task
@valarLip
5
fix fp8 blockscale configs for deepseek
#1232 opened Oct 21, 2025 by juuso-oskari Loading…
@juuso-oskari @anhminhnguyenhoang
1
Enable custom op and avoid graph breaks (#740)
#1228 opened Oct 21, 2025 by divakar-amd Loading…
1
[TRITON] Gluon softmax implementation
#1227 opened Oct 20, 2025 by eky-amd • Draft
add tune file for moe ops in deepseek
#1222 opened Oct 19, 2025 by PerryZhang01 Loading…
1 task
fix repeated unnecessary device check
#1221 opened Oct 18, 2025 by b8zhong Loading…
1 task done
11
CI: Optimize the Triton setup process
#1213 opened Oct 17, 2025 by gyohuangxin • Draft
[CK_TILE] fmha: Add backward pass support for padded inputs
#1212 opened Oct 17, 2025 by Jeff-Huang Loading…
1 task
Update build_triton.sh to add timeout and retry
#1211 opened Oct 17, 2025 by gyohuangxin • Draft
[TRITON] Add AOT compiled FP4 weight preshuffle kernels 355 wip
#1209 opened Oct 17, 2025 by azaidy Loading…
1
CI: [MI35x] Add ROCm 6.4.2 tests for standard and multigpu tests
#1208 opened Oct 17, 2025 by gyohuangxin • Draft
update mi308 fmoe fp16 asm,MI35x
#1201 opened Oct 15, 2025 by amd-ruitang3 Loading…
1 task
1
[Triton] A8W8 blockscale GEMM tuning for Qwen3
#1195 opened Oct 14, 2025 by anhminhnguyenhoang • Draft
1 task
@anhminhnguyenhoang
CI: Move triton tests from mi325 to mi355
#1183 opened Oct 13, 2025 by gyohuangxin Loading…
1
ProTip! Add no:assignee to see everything that’s not assigned.