Utility for Multi-GPU Node Configuration

rghosh81 pts0 comments

Release v0.1.0 — NVSwitch driver + Fabric Manager installer · rghosh08/nvidia-nvswitch-setup · GitHub

//releases/show" data-turbo-transient="true" />

Skip to content

Search or jump to...

Search code, repositories, users, issues, pull requests...

-->

Search

Clear

Search syntax tips

Provide feedback

--><br>We read every piece of feedback, and take your input very seriously.

Include my email address so I can be contacted

Cancel

Submit feedback

Saved searches

Use saved searches to filter your results more quickly

-->

Name

Query

To see all available qualifiers, see our documentation.

Cancel

Create saved search

Sign in

//releases/show;ref_cta:Sign up;ref_loc:header logged out"}"<br>Sign up

Appearance settings

Resetting focus

You signed in with another tab or window. Reload to refresh your session.<br>You signed out in another tab or window. Reload to refresh your session.<br>You switched accounts on another tab or window. Reload to refresh your session.

Dismiss alert

{{ message }}

rghosh08

nvidia-nvswitch-setup

Public

Notifications<br>You must be signed in to change notification settings

Fork

Star

v0.1.0 — NVSwitch driver + Fabric Manager installer

Latest

Latest

Compare

Choose a tag to compare

Sorry, something went wrong.

Filter

Loading

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

No results found

View all tags

rghosh08

released this

26 Jun 18:19

v0.1.0

cfbbd05

First release. A small, idempotent installer for headless multi-GPU NVSwitch boxes (A100/H100 SXM, e.g. AWS p4d/p4de/p5) on stock Ubuntu.

Install

sudo apt-get install -y ./nvidia-nvswitch-setup_0.1.0_all.deb<br>sudo nvidia-nvswitch-setup # auto-pick newest matched branch<br>sudo nvidia-nvswitch-setup --dry-run # show the plan, change nothing

What it does

Auto-selects the newest driver branch whose server-open driver and nvidia-fabricmanager versions match exactly (FM refuses to start otherwise).

Installs the lean headless server-open set (no X11).

Verifies the DKMS kernel module actually built for the running kernel before declaring success — so you don't get a "successful" install with no usable GPUs.

Starts Fabric Manager (required to initialize the NVSwitch fabric).

No driver version is pinned, so the package does not rot as new branches ship.

Verified on an 8× A100-SXM4-80GB host (kernel 6.17, driver 595.71.05).

Assets

Loading

Uh oh!

There was an error while loading. Please reload this page.

-->

All reactions

You can’t perform that action at this time.

nvswitch driver nvidia search reload fabric

Related Articles