[PR #306] [MERGED] Optimize `cached_notify` by TMA #512

New issue

Closed

opened 2026-10-27 21:03:29 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-10-27 21:03:29 +00:00

Owner

📋 Pull Request Information

Original PR: https://github.com/deepseek-ai/DeepEP/pull/306
Author: @sphish
Created: 7/15/2025
Status: ✅ Merged
Merged: 7/16/2025
Merged by: @LyricZhao

Base: main ← Head: opt-combine-notify

📝 Commits (4)

e5f9cb1 Fix rdma head movement
008e399 Optimize cached_notify by using TMA.
772b735 Fix
a1fcfea Small fix

📊 Changes

1 file changed (+50 additions, -10 deletions)

View changed files

📝 csrc/kernels/internode.cu (+50 -10)

📄 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/306 **Author:** [@sphish](https://github.com/sphish) **Created:** 7/15/2025 **Status:** ✅ Merged **Merged:** 7/16/2025 **Merged by:** [@LyricZhao](https://github.com/LyricZhao) **Base:** `main` ← **Head:** `opt-combine-notify` --- ### 📝 Commits (4) - [`e5f9cb1`](https://github.com/deepseek-ai/DeepEP/commit/e5f9cb13f122731832949920141293c39e96e737) Fix rdma head movement - [`008e399`](https://github.com/deepseek-ai/DeepEP/commit/008e399eb01b746439c0df88fa78e68eb3f3ec3c) Optimize `cached_notify` by using TMA. - [`772b735`](https://github.com/deepseek-ai/DeepEP/commit/772b7357c58b3f8c94e608083b9d8ce9fcbe4772) Fix - [`a1fcfea`](https://github.com/deepseek-ai/DeepEP/commit/a1fcfea47eab62a6969dbd575ee0b66417de5bc3) Small fix ### 📊 Changes **1 file changed** (+50 additions, -10 deletions) <details> <summary>View changed files</summary> 📝 `csrc/kernels/internode.cu` (+50 -10) </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

2026-10-27 21:03:29 +00:00