Skip to content

Pull requests: deepseek-ai/DeepEP

New pull request New
26 Open 127 Closed
26 Open 127 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

Higher Performance with Lower SM Occupancy through Zero-Copy and TMA Offloading
#453 opened Oct 14, 2025 by monethuang1 Loading…
Decouple the dispatch and buffer allocation on hybrid-ep
#448 opened Oct 10, 2025 by Autumn1998 Loading…
docs(readme): fix typos, clarify wording
#444 opened Oct 9, 2025 by Dhie-boop Loading…
Fix racing condition in large batch size
#440 opened Oct 2, 2025 by fzyzcjy Loading…
Support CUDA Graph for internode dispatch normal kernel
#438 opened Sep 30, 2025 by yifeizhang-c Loading…
1
Fix get_nvl_buffer_size_hint and get_rdma_buffer_size_hint
#434 opened Sep 26, 2025 by yuantailing Loading…
opt ll dispatch layered algo
#425 opened Sep 24, 2025 by alpha-baby Loading…
Support per tensor transfer
#416 opened Sep 18, 2025 by ayrnb Loading…
atomic_clean_flag in low latency combine, seems useless ?
#409 opened Sep 16, 2025 by zhoutianzi666 Loading…
3
One branch that contains several optimizations and features
#405 opened Sep 15, 2025 by fzyzcjy • Draft
Add imbalance factor in test_low_latency
#393 opened Sep 4, 2025 by JianboDong Loading…
1
[Feat] Single Batch Overlap (SBO): Overlaping of Down GEMM with Combine Send
#390 opened Sep 2, 2025 by Zqy11 Loading…
10
Another computation communication overlap
#388 opened Sep 1, 2025 by fzyzcjy • Draft
delete some redundant logic using cg::this_grid().sync();
#357 opened Aug 9, 2025 by zhoutianzi666 Loading…
3
Feature/sm free normal kernel
#347 opened Jul 31, 2025 by ZhiyiHu1999 Loading…
4 tasks
2
Support nvfp4 low latency mode dispatch
#341 opened Jul 30, 2025 by shifangx Loading…
11
feat(buffer): implement dynamic buffer resizing for simplicity
#340 opened Jul 30, 2025 by MengAiDev Loading…
Support prefill with 2 GPUs
#331 opened Jul 27, 2025 by fzyzcjy Loading…
enhance warp copy efficiency in cached_notify()
#315 opened Jul 18, 2025 by ZhiyiHu1999 Loading…
2
support low latency dispatch tma
#293 opened Jul 10, 2025 by ayrnb Loading…
2
Tiny support custom nvcc flags
#280 opened Jul 5, 2025 by fzyzcjy Loading…
Allow using few SMs for low-latency mode
#277 opened Jul 3, 2025 by fzyzcjy Loading…
3
Computation communication overlap
#249 opened Jun 24, 2025 by fzyzcjy • Draft
Optimize low latency combine recv kernel (about 3.0x speedup)
#248 opened Jun 23, 2025 by fzyzcjy Loading…
7
Support other NVLink scenarios
#218 opened Jun 17, 2025 by fzyzcjy Loading…
11
ProTip! Mix and match filters to narrow down what you’re looking for.