Skip to content


This page includes information about the self-hosted runners that are available from NVIDIA and how to use them.

CPU Runners

The CPU labeled runners are backed by various EC2 instances and do not have any GPUs installed.

Label EC2 Machine Type
linux-amd64-cpu4 m5d.xlarge (specs)
linux-amd64-cpu8 m5d.2xlarge (specs)
linux-amd64-cpu16 m5d.4xlarge (specs)
linux-arm64-cpu4 m6gd.xlarge (specs)
linux-arm64-cpu8 m6gd.2xlarge (specs)
linux-arm64-cpu16 m6gd.4xlarge (specs)


The CPU label names consist of the following components:

^     ^     ^  ^
|     |     |  |
|     |     |  CPU Core Count
|     |     CPU Designator
|     Architecture
Operating System

GPU Runners

The GPU labeled runners have the GPUs specified in the table below installed.


GPU jobs have two requirements. If these requirements aren't met, the GitHub Actions job will fail. See the Example section below for an example.

  1. They must run in a container (i.e. nvidia/cuda:12.0.0-base-ubuntu22.04)
  2. They must set the NVIDIA_VISIBLE_DEVICES: ${{ env.NVIDIA_VISIBLE_DEVICES }} container environment variable
Label GPU Driver Version # of GPUs
linux-amd64-gpu-v100-earliest-1 V100 470 1
linux-amd64-gpu-v100-latest-1 V100 525 1
linux-arm64-gpu-a100-latest-1 A100 525 1
linux-amd64-gpu-t4-latest-1 T4 525 1


The GPU label names consist of the following components:

^     ^     ^   ^  ^      ^
|     |     |   |  |      |
|     |     |   |  |      Number of GPUs Available
|     |     |   |  GPU Driver Version
|     |     |   GPU Type
|     |     GPU Designator
|     Architecture
Operating System

Due to limited GPU capacity and the overhead associated with manually rotating self-hosted runner labels when GPU drivers are updated, there are no driver-specific self-hosted runner labels (e.g. linux-amd64-gpu-t4-525-1).

Instead, the driver-version designators earliest and latest are used. The values of these designators represent the oldest and newest NVIDIA supported drivers at any given time.

The chart above will be kept up-to-date with the corresponding driver versions at any given time.

Supported organizations will be notified whenever these versions are scheduled to be updated.


The code snippet below shows how the labels above may be utilized in a GitHub Action workflow.

name: Test Self Hosted Runners
on: push
    runs-on: linux-amd64-cpu8
      - name: hello
        run: echo "hello"
    runs-on: linux-amd64-gpu-v100-latest-1
    container: # GPU jobs must run in a container
      image: nvidia/cuda:12.0.0-base-ubuntu22.04
        NVIDIA_VISIBLE_DEVICES: ${{ env.NVIDIA_VISIBLE_DEVICES }} # GPU jobs must set this container env variable
      - name: hello
        run: |
          echo "hello"

For additional details on self-hosted runner usage, see the official GitHub Action documentation page here.