[PR #83] [MERGED] Use 1D TMA store instead of 3D #238

New issue

Closed

opened 2026-10-26 21:15:12 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-10-26 21:15:12 +00:00

Owner

📋 Pull Request Information

Original PR: https://github.com/deepseek-ai/DeepGEMM/pull/83
Author: @LyricZhao
Created: 4/11/2025
Status: ✅ Merged
Merged: 4/11/2025
Merged by: @LyricZhao

Base: main ← Head: tma-1d-store

📝 Commits (4)

a77009c Make partition pipelined
8041ed7 Use 1D TMA store
99eb6ec Remove useless STSM
b0d6481 OOB bugs fixed

📊 Changes

5 files changed (+24 additions, -80 deletions)

View changed files

📝 deep_gemm/include/deep_gemm/fp8_gemm.cuh (+21 -53)
📝 deep_gemm/include/deep_gemm/tma_utils.cuh (+0 -21)
📝 deep_gemm/jit_kernels/gemm.py (+1 -2)
📝 deep_gemm/jit_kernels/m_grouped_gemm.py (+1 -2)
📝 indexing/main.cu (+1 -2)

📄 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/DeepGEMM/pull/83 **Author:** [@LyricZhao](https://github.com/LyricZhao) **Created:** 4/11/2025 **Status:** ✅ Merged **Merged:** 4/11/2025 **Merged by:** [@LyricZhao](https://github.com/LyricZhao) **Base:** `main` ← **Head:** `tma-1d-store` --- ### 📝 Commits (4) - [`a77009c`](https://github.com/deepseek-ai/DeepGEMM/commit/a77009cb149b132eebe3d8f0a11032c2ad2db8fb) Make partition pipelined - [`8041ed7`](https://github.com/deepseek-ai/DeepGEMM/commit/8041ed71642e58e8a36d63d3a522e6e7f218eeeb) Use 1D TMA store - [`99eb6ec`](https://github.com/deepseek-ai/DeepGEMM/commit/99eb6ec563807cb8879633281e01017897a1c372) Remove useless STSM - [`b0d6481`](https://github.com/deepseek-ai/DeepGEMM/commit/b0d64817a77d93fe07ec22aee9cd877b3ce538cc) OOB bugs fixed ### 📊 Changes **5 files changed** (+24 additions, -80 deletions) <details> <summary>View changed files</summary> 📝 `deep_gemm/include/deep_gemm/fp8_gemm.cuh` (+21 -53) 📝 `deep_gemm/include/deep_gemm/tma_utils.cuh` (+0 -21) 📝 `deep_gemm/jit_kernels/gemm.py` (+1 -2) 📝 `deep_gemm/jit_kernels/m_grouped_gemm.py` (+1 -2) 📝 `indexing/main.cu` (+1 -2) </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 21:15:12 +00:00