Skip to content

Pull requests: deepseek-ai/DeepGEMM

New pull request New
33 Open 156 Closed
33 Open 156 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: sm90 support mxfp8 fp8 groupgemm
#362 opened Jun 16, 2026 by zhangxiaolei123456 Loading…
fix: index CUDA sources/headers when generating .pyi stubs
#361 opened Jun 14, 2026 by Osamaali313 Loading…
2
Add SM90 MegaMoE by pingpong and cooperative
#360 opened Jun 13, 2026 by mpdfdfl Loading…
Draft: Add green-context split-kernel MegaMoE features
#357 opened Jun 12, 2026 by RayWang96 Collaborator Loading…
fix: use importlib to load scripts/generate_pyi.py instead of package…
#355 opened Jun 5, 2026 by Mikezhang001 Loading…
Add SM90 FP8 MegaMoE split kernel
#352 opened Jun 3, 2026 by AichenF Loading…
5
Add shape-aware recommended alignment for SM90 small-M grouped GEMM
#350 opened Jun 2, 2026 by qescccczmr Loading…
feat: support mxfp4 × mxfp4 gemm on sm100
#348 opened Jun 1, 2026 by z52527 Loading…
Fix a race condition in contiguous k-grouped GEMM where in-flight tensormaps are updated in-place
#343 opened May 29, 2026 by dfyz Loading…
12
Fix TMEM lane address for debug mode across all SM100 kernels
#341 opened May 28, 2026 by yejunjin Loading…
Add SM90 FP8 paged MQA logits support for next_n=3Fix sm90 nextn3 paged mqa logits
#340 opened May 28, 2026 by yangsiqt Loading…
Fix ue8m0 packing: mask mantissa bits when extracting fp32 exponents
#337 opened May 19, 2026 by yhyang201 Loading…
feat: support sm90_fp8_fp4 kernel
#332 opened May 11, 2026 by zhangxiaolei123456 Loading…
fix: use 64-bit mega moe sf offsets
#331 opened May 10, 2026 by netaddi Loading…
fix: replace numeric_limits::infinity() with literal to fix CUDA 12.8 NVRTC compilation
#327 opened May 5, 2026 by kuishou68 Loading…
1
[WIP] support sm90_fp4_mqa_logits
#326 opened May 1, 2026 by laixinn • Draft
feat: add sm120 support for DeepGEMM
#324 opened May 1, 2026 by leavelet Loading…
13
feat: MegaMOE adaptation for SM90
#323 opened Apr 30, 2026 by qiushixiaoyu Loading…
28
Support SM120
#318 opened Apr 25, 2026 by jasl Contributor Loading…
17
fix: correct operator precedence in pack_ue8m0_to_int assertion
#310 opened Apr 19, 2026 by kuishou68 Loading…
1
Fix JIT cache race condition with multi-process compilation
#302 opened Apr 11, 2026 by Gregory-Pereira Loading…
3
feat: support bf16 output and plain TMA writes in k_grouped_gemm on SM90;
#298 opened Mar 26, 2026 by fedorovgv Loading…
fix: No module named 'scripts.generate_pyi'
#288 opened Feb 22, 2026 by HollowMan6 Loading…
4
feat: add w4afp8 on Hopper GPUs
#287 opened Feb 9, 2026 by foreverrookie Loading…
Windows support
#284 opened Feb 2, 2026 by tori29umai0123 Loading…
ProTip! Type g i on any issue or pull request to go back to the issue listing page.