Skip to content

Add NCCL_NCHANNELS_PER_NET_PEER to NCCL documentation #295

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Nov 5, 2025
Merged

Add NCCL_NCHANNELS_PER_NET_PEER to NCCL documentation #295

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
f6587ec
Add NCCL_NCHANNELS_PER_NET_PEER to NCCL documentation
msimberg Nov 5, 2025
4d95323
Update NCCL_NCHANNELS_PER_NET_PEER section and add "recv" to whitelist
msimberg Nov 5, 2025
bea05e1
Refer to paper for NCCL_NCHANNELS_PER_NET_PEER
msimberg Nov 5, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions .github/actions/spelling/allow.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -273,6 +273,7 @@ pytorch
quantumespresso
quasiparticles
quickstart
recv
rgw
ripgrep
rocm
Expand Down
10 changes: 10 additions & 0 deletions docs/software/communication/nccl.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,6 +22,16 @@ While the container engine sets these automatically when using the NCCL hook, th

[_Demystifying NCCL: An In-depth Analysis of GPU Communication Protocols and Algorithms_](https://arxiv.org/abs/2507.04786v2) contains detailed information about NCCL algorithms and protocols, which can be helpful for deciding if your application could benefit from an alternative configuration.

In addition to the above variables, setting `NCCL_NCHANNELS_PER_NET_PEER` can improve point-to-point performance (operations based directly on send/recv):

```bash
export NCCL_NCHANNELS_PER_NET_PEER=4
```

A value of 4 is generally a good compromise to improve point-to-point performance without affecting collectives performance.
Setting it to a higher value such as 16 or 32 can still further improve send/recv performance, but may degrade collectives performance, so the optimal value depends on the mix of operations used in an application.
The option is undocumented, but [this issue](https://github.com/NVIDIA/nccl/issues/1272) and the paper linked above contain additional details.

!!! warning "NCCL watchdog timeout or hanging process"
In some cases, still under investigation, NCCL may hang resulting in a stuck process or a watchdog timeout error.
In this scenario, we recommend disabling Slingshot eager messages with the following workaround:
Expand Down