deepseek-ai/DeepEP
1
0
Fork 0
mirror of https://github.com/deepseek-ai/DeepEP.git synced 2026-11-05 00:47:32 +00:00
Code Issues 308 Projects Packages Wiki Activity

[GH-ISSUE #77] cuda failed with too many blocks in cooperative launch in H20 #72

New issue
Closed
opened 2026-10-26 09:53:45 +00:00 by bsakaguchi · 1 comment
bsakaguchi commented 2026-10-26 09:53:45 +00:00
Owner
Copy link

Originally created by @cll24 on GitHub (Mar 16, 2025).
Original GitHub issue: https://github.com/deepseek-ai/DeepEP/issues/77

I met the following RuntimeError when I was running the python tests/test_low_latency.py in H20 with 4 RoCE NICs. Does anyone met this error before, and would you please share your method of solving this problem. Here is output of nvidia-smi and ibv_devinfo

Image

Image

Image

Originally created by @cll24 on GitHub (Mar 16, 2025). Original GitHub issue: https://github.com/deepseek-ai/DeepEP/issues/77 I met the following RuntimeError when I was running the `python tests/test_low_latency.py` in H20 with 4 RoCE NICs. Does anyone met this error before, and would you please share your method of solving this problem. Here is output of nvidia-smi and ibv_devinfo ![Image](https://github.com/user-attachments/assets/762bfe01-0f8c-45f7-a7eb-2e638907233c) ![Image](https://github.com/user-attachments/assets/421429bd-2bc3-499a-b872-de291794cd5d) ![Image](https://github.com/user-attachments/assets/3ff2e607-bd94-4aab-979e-501f696d6fc5)
bsakaguchi commented 2026-10-26 09:53:46 +00:00
Author
Owner
Copy link

@LyricZhao commented on GitHub (Mar 17, 2025):

Duplicated issue, please see and reply in #15, feel free to reopen this issue if your problem is not the same.

<!-- gh-comment-id:2727790558 --> @LyricZhao commented on GitHub (Mar 17, 2025): Duplicated issue, please see and reply in #15, feel free to reopen this issue if your problem is not the same.
Sign in to join this conversation.
No Branch/Tag specified
Branches Tags
main
hybrid-ep
chore/remove-clang-format-yapf
v2.5
nvDev
try-fix-cu128
mori-ep
antgroup-opt
compat-old-torch
add-linter
eager-rdma
v1.2.1
Labels
Clear labels   
pull-request
Mirrored from GitHub Pull Request

No labels
pull-request
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
deepseek-ai/DeepEP#72
No description provided.