[PR #738] fix: support system NCCL installations #741

New issue

Open

opened 2026-10-27 21:05:48 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-10-27 21:05:48 +00:00

Owner

📋 Pull Request Information

Original PR: https://github.com/deepseek-ai/DeepEP/pull/738
Author: @janbernloehr
Created: 8/25/2026
Status: 🔄 Open

Base: main ← Head: system-nccl-main

📝 Commits (1)

7014299 fix: support system NCCL installations

📊 Changes

2 files changed (+77 additions, -9 deletions)

View changed files

📝 deep_ep/__init__.py (+32 -4)
📝 setup.py (+45 -5)

📄 Description

Summary

DeepEP currently assumes NCCL is installed through the PyPI package
nvidia-nccl-cu12. This fails for standard Ubuntu installations where NCCL
is provided externally, for example:

/usr/include/nccl.h
/usr/lib/aarch64-linux-gnu/libnccl.so

This change:

Discovers NCCL through standard multiarch/system library locations.
Preserves the existing PyPI-package discovery path.
Links the exact discovered library without overriding the system NCCL version.
Avoids adding an unnecessary RPATH for system-installed NCCL.
Uses the same discovery logic for runtime checks and JIT compilation.

HybridEP does not require this change: its NIXL path skips NCCL/DOCA, while its
DOCA path intentionally builds the bundled NCCL submodule components.

Testing (internal)

Python syntax and whitespace checks passed.
System NCCL discovery was validated in the Core Models mbridge build.
DeepEP 01dc3aaac82068020353dce2c302e38153c0bfaa built and installed
successfully with system NCCL.
deep_ep._C metadata validation and import deep_ep passed.

_{🔄 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/738 **Author:** [@janbernloehr](https://github.com/janbernloehr) **Created:** 8/25/2026 **Status:** 🔄 Open **Base:** `main` ← **Head:** `system-nccl-main` --- ### 📝 Commits (1) - [`7014299`](https://github.com/deepseek-ai/DeepEP/commit/70142994fe2be754cfa35d92ff1b34f0e228ec35) fix: support system NCCL installations ### 📊 Changes **2 files changed** (+77 additions, -9 deletions) <details> <summary>View changed files</summary> 📝 `deep_ep/__init__.py` (+32 -4) 📝 `setup.py` (+45 -5) </details> ### 📄 Description ## Summary DeepEP currently assumes NCCL is installed through the PyPI package `nvidia-nccl-cu12`. This fails for standard Ubuntu installations where NCCL is provided externally, for example: - `/usr/include/nccl.h` - `/usr/lib/aarch64-linux-gnu/libnccl.so` This change: - Discovers NCCL through standard multiarch/system library locations. - Preserves the existing PyPI-package discovery path. - Links the exact discovered library without overriding the system NCCL version. - Avoids adding an unnecessary RPATH for system-installed NCCL. - Uses the same discovery logic for runtime checks and JIT compilation. HybridEP does not require this change: its NIXL path skips NCCL/DOCA, while its DOCA path intentionally builds the bundled NCCL submodule components. ## Testing (internal) - Python syntax and whitespace checks passed. - System NCCL discovery was validated in the Core Models mbridge build. - DeepEP `01dc3aaac82068020353dce2c302e38153c0bfaa` built and installed successfully with system NCCL. - `deep_ep._C` metadata validation and `import deep_ep` passed. --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

bsakaguchi added the

pull-request

label

2026-10-27 21:05:48 +00:00

bsakaguchi referenced this issue

2026-10-27 21:05:50 +00:00

[PR #741] docs: point Development section to develop.sh (fix stale ln -s for V2) #746