Skip to content

Pull requests: ROCm/aiter

New pull request New
490 Open 3,717 Closed
490 Open 3,717 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

feat(triton): enable gfx1100 MXFP4 MoE
#4640 opened Aug 8, 2026 by skyguan92 • Draft
5 tasks done
1
[Bugfix] Fix two build failures: hipify qualifier loss and std::min
#4639 opened Aug 8, 2026 by echen4096 Loading…
1 task
1
fix(quant): round scaled int8 casts to nearest
#4637 opened Aug 7, 2026 by skyguan92 Loading…
1
Tune mla_decode_rope fp32 config for gfx950 ci:triton-300x ci:triton-355
#4632 opened Aug 7, 2026 by nidal567 Contributor Loading…
1 task done
@nidal567
2
unify the loading and add a UT for configs ci:triton-300x
#4630 opened Aug 7, 2026 by Boss2002n Contributor • 2/2 Loading…
1 task
7
[GLM-5.2 MXFP4][Tuning] Retune MXFP4 fused-MoE for gfx950 (TP4 + TP8)
#4629 opened Aug 7, 2026 by nholmber Contributor Loading…
1 task done
1
fix(fused_moe): restore gfx942 FLAT kernel support
#4628 opened Aug 7, 2026 by bingxche Contributor Loading…
1 of 4 tasks
3
new MHA v4 entrypoint ; hosts a spectrum of quantized & sparse attention kernels
#4627 opened Aug 7, 2026 by jcaraban Contributor Loading…
1 task done
@juuso-oskari @Chi-Chu319
9
Gfx1250 flydsl batched gemm
#4626 opened Aug 7, 2026 by XingerZhu Contributor Loading…
1 task
6
[MLA] Support 96-head 128-dim reduction
#4625 opened Aug 7, 2026 by LiuYinfeng01 Contributor Loading…
1
CI: resolve container Python for release builds
#4624 opened Aug 7, 2026 by gyohuangxin Member • Draft
3
[FlyDSL] Replace the split-K atomic combine with a workspace + reduce kernel
#4622 opened Aug 7, 2026 by JohnQinAMD Contributor Loading…
6
fix: car graph mode err when pytorch set expandable_segments:True
#4621 opened Aug 7, 2026 by TennyWang1223 Contributor Loading…
1 task done
1
2
[MoE] Added Gelu with tanh approx for CK XDL 2-stage MoE
#4620 opened Aug 7, 2026 by a-sidorova Contributor Loading…
4
CI: run vLLM disagg from upstream main ci:vllm-di Trigger vLLM disaggregated inference CI
#4619 opened Aug 7, 2026 by JiaoliangYu Contributor Loading…
1
[FlyDSL]optimize mxfp4 moe stage2 perf for gfx950
#4618 opened Aug 7, 2026 by binding7012 Loading…
1
feat(fused_moe): accept a caller-provided output buffer
#4617 opened Aug 6, 2026 by RolaoDenthu Contributor Loading…
1 task
1
MLA kernel flydsl bf16
#4616 opened Aug 6, 2026 by ahmed-bsod Contributor Loading…
8
[Draft] Register FlyDSL operations in torch
#4615 opened Aug 6, 2026 by mohbasit Loading…
1 task
[TRITON][GLUON][GFX950] Add Unified Attention Gluon Kernel
#4614 opened Aug 6, 2026 by cagrikymk Contributor Loading…
2 tasks done
3
[Triton/Gluon] Add dytpe dirs to the configs
#4613 opened Aug 6, 2026 by Boss2002n Contributor • 1/2 Loading…
1
[BUG][CK][MHA] Fix for MHA with softmax-sink
#4612 opened Aug 6, 2026 by shurale-nkn Loading…
1 task
2
chore(aot/flydsl): drop per-kernel "[OK] compile" log; silence false-positive mxmoe warning
#4611 opened Aug 6, 2026 by zhiding512 Contributor Loading…
2
Flydsl attn res
#4610 opened Aug 6, 2026 by anhminhnguyenhoang Contributor • Draft
1 task
1
[gfx1250] Fuse A4W4 stage1 FP4 quantization
#4607 opened Aug 6, 2026 by XiaobingSuper Contributor Loading…
5 tasks done
3
ProTip! no:milestone will show everything without a milestone.