mirror of
https://github.com/deepseek-ai/DeepEP.git
synced 2026-11-05 00:47:32 +00:00
[PR #204] [MERGED] Support Ampere architecture #462
Labels
No labels
pull-request
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
deepseek-ai/DeepEP#462
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
📋 Pull Request Information
Original PR: https://github.com/deepseek-ai/DeepEP/pull/204
Author: @LyricZhao
Created: 6/11/2025
Status: ✅ Merged
Merged: 6/11/2025
Merged by: @LyricZhao
Base:
main← Head:ampere-support📝 Commits (10+)
d5664f3Update README5e35592Updatesetup.pyd3ae62dFix headers36fac36AddDISABLE_NVSHMEMfor APIs7635af2Fix launch910c338Fix TMA settings422e2a3Fix TMA usagesd2ca520Fix dlink1204a56Separate layout kernels4664e83Update version📊 Changes
16 files changed (+413 additions, -174 deletions)
View changed files
📝
README.md(+16 -4)📝
csrc/config.hpp(+6 -0)📝
csrc/deep_ep.cpp(+63 -9)📝
csrc/kernels/CMakeLists.txt(+3 -2)📝
csrc/kernels/api.cuh(+11 -6)📝
csrc/kernels/configs.cuh(+15 -1)📝
csrc/kernels/internode.cu(+0 -125)📝
csrc/kernels/intranode.cu(+25 -4)📝
csrc/kernels/launch.cuh(+31 -0)➕
csrc/kernels/layout.cu(+136 -0)📝
csrc/kernels/runtime.cu(+5 -0)📝
csrc/kernels/utils.cuh(+5 -0)📝
deep_ep/buffer.py(+6 -1)📝
deep_ep/utils.py(+28 -0)📝
setup.py(+58 -16)📝
tests/test_intranode.py(+5 -6)📄 Description
Please notice that this support is for intranode only.
🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.