[PR #73] [MERGED] Low latency kernels use rdma atomic to support AR. #431

New issue

Closed

opened 2026-10-26 10:00:40 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-10-26 10:00:40 +00:00

Owner

📋 Pull Request Information

Original PR: https://github.com/deepseek-ai/DeepEP/pull/73
Author: @sphish
Created: 3/14/2025
Status: ✅ Merged
Merged: 3/14/2025
Merged by: @LyricZhao

Base: main ← Head: p2p-signal

📝 Commits (2)

2d0cf41 Low latency kernels use rdma atomic to support AR.
38cdaf3 Fix style.

📊 Changes

3 files changed (+69 additions, -102 deletions)

View changed files

📝 csrc/kernels/ibgda_device.cuh (+63 -56)
📝 csrc/kernels/internode_ll.cu (+5 -21)
📝 csrc/kernels/runtime.cu (+1 -25)

📄 Description

No description provided

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/deepseek-ai/DeepEP/pull/73 **Author:** [@sphish](https://github.com/sphish) **Created:** 3/14/2025 **Status:** ✅ Merged **Merged:** 3/14/2025 **Merged by:** [@LyricZhao](https://github.com/LyricZhao) **Base:** `main` ← **Head:** `p2p-signal` --- ### 📝 Commits (2) - [`2d0cf41`](https://github.com/deepseek-ai/DeepEP/commit/2d0cf41dd1900b105d74cb071f4cac35e3fb6f47) Low latency kernels use rdma atomic to support AR. - [`38cdaf3`](https://github.com/deepseek-ai/DeepEP/commit/38cdaf390c202f6c7137b0ef7e6d3edc21d27161) Fix style. ### 📊 Changes **3 files changed** (+69 additions, -102 deletions) <details> <summary>View changed files</summary> 📝 `csrc/kernels/ibgda_device.cuh` (+63 -56) 📝 `csrc/kernels/internode_ll.cu` (+5 -21) 📝 `csrc/kernels/runtime.cu` (+1 -25) </details> ### 📄 Description _No description provided_ --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

bsakaguchi

2026-10-26 10:00:40 +00:00