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 #488] Question About num_rc_per_pe in Normal Dispatch #316

New issue
Closed
opened 2026-10-26 09:58:30 +00:00 by bsakaguchi · 1 comment
bsakaguchi commented 2026-10-26 09:58:30 +00:00
Owner
Copy link

Originally created by @michaelchen1996 on GitHub (Nov 10, 2025).
Original GitHub issue: https://github.com/deepseek-ai/DeepEP/issues/488

https://github.com/deepseek-ai/DeepEP/pull/181 fixed a timeout problem by forbidding concurrent submissions of QP.

However, in normal dispatch

EP_DEVICE_ASSERT(ibgda_get_state()->num_rc_per_pe == num_channels or ibgda_get_state()->num_rc_per_pe >= num_sms);

Why is num_rc_per_pe == num_channels still allowed?

Thank you!

Originally created by @michaelchen1996 on GitHub (Nov 10, 2025). Original GitHub issue: https://github.com/deepseek-ai/DeepEP/issues/488 https://github.com/deepseek-ai/DeepEP/pull/181 fixed a timeout problem by forbidding concurrent submissions of QP. However, in normal dispatch ``` EP_DEVICE_ASSERT(ibgda_get_state()->num_rc_per_pe == num_channels or ibgda_get_state()->num_rc_per_pe >= num_sms); ``` Why is `num_rc_per_pe == num_channels` still allowed? Thank you!
bsakaguchi commented 2026-10-26 09:58:31 +00:00
Author
Owner
Copy link

@sphish commented on GitHub (Nov 12, 2025):

This PR does not actually fix the underlying issue; it only increases the number of QPs, which leads to better performance. The original QP configuration is still functional, and to maintain forward compatibility, we have kept the original QP settings in the assertions.

<!-- gh-comment-id:3519700919 --> @sphish commented on GitHub (Nov 12, 2025): This PR does not actually fix the underlying issue; it only increases the number of QPs, which leads to better performance. The original QP configuration is still functional, and to maintain forward compatibility, we have kept the original QP settings in the assertions.
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#316
No description provided.