centersl/docker-docs
1
0
Fork 0
mirror of https://github.com/docker/docs.git synced 2026-03-27 06:18:55 +07:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #24433 from doringeman/containers-gpu

Browse Source 
docs(containers): move GPU access to dedicated page
This commit is contained in:
David Karlsson
2026-03-20 07:01:23 +01:00
committed by GitHub
parent dd599144b1 0913862a02
commit 55e175d1a7
2 changed files with 116 additions and 83 deletions
Download Patch File Download Diff File Expand all files Collapse all files

112
content/manuals/engine/containers/gpu.md Normal file
View File

@@ -0,0 +1,112 @@
---
title: GPU access
weight: 40
description: How to access NVIDIA GPUs from a container
keywords: docker, GPU, NVIDIA, cuda, nvidia-smi, device, container toolkit
---
## Access an NVIDIA GPU
### Prerequisites
Visit the official [NVIDIA drivers page](https://www.nvidia.com/Download/index.aspx)
to download and install the proper drivers. Reboot your system once you have
done so.
Verify that your GPU is running and accessible.
### Install NVIDIA Container Toolkit
Follow the official NVIDIA Container Toolkit [installation instructions](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html).
### Expose GPUs for use
Include the `--gpus` flag when you start a container to access GPU resources.
To expose all available GPUs:
```console
$ docker run -it --rm --gpus all ubuntu nvidia-smi
```
The output looks like the following:
```text
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.288.01 Driver Version: 535.288.01 CUDA Version: 12.2 |
|-----------------------------------------+----------------------+----------------------+
| GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+======================+======================|
| 0 NVIDIA L4 Off | 00000000:31:00.0 Off | 0 |
| N/A 40C P0 27W / 72W | 0MiB / 23034MiB | 4% Default |
| | | N/A |
+-----------------------------------------+----------------------+----------------------+
+---------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=======================================================================================|
| No running processes found |
+---------------------------------------------------------------------------------------+
```
The leftmost column in the GPU table shows the index of each GPU (`0` for the
NVIDIA L4 in the previous example). Use these index numbers to target specific GPUs
with the `device` option.
To expose a single GPU by index:
```console
$ docker run -it --rm --gpus device=0 ubuntu nvidia-smi
```
To expose a GPU by its UUID, first list UUIDs with `nvidia-smi -L`:
```console
$ nvidia-smi -L
GPU 0: NVIDIA L4 (UUID: GPU-3a23c669-1f69-c64e-cf85-44e9b07e7a2a)
```
Then pass the UUID to `--gpus`:
```console
$ docker run -it --rm --gpus device=GPU-3a23c669-1f69-c64e-cf85-44e9b07e7a2a ubuntu nvidia-smi
```
On systems with multiple GPUs, you can expose several by index. The `device`
value must be quoted because it contains a comma:
```console
$ docker run -it --rm --gpus '"device=0,2"' ubuntu nvidia-smi
```
This exposes the GPUs at index `0` and `2` — the first and third GPUs listed in
`nvidia-smi` output.
> [!NOTE]
>
> NVIDIA GPUs can only be accessed by systems running a single engine.
### Set NVIDIA capabilities
You can set capabilities manually. For example, on Ubuntu you can run the
following:
```console
$ docker run --gpus 'all,capabilities=utility' --rm ubuntu nvidia-smi
```
This enables the `utility` driver capability which adds the `nvidia-smi` tool to
the container.
Capabilities and other configurations can be set in images via environment
variables. For valid variables, see the
[nvidia-container-toolkit](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/docker-specialized.html)
documentation. These variables can be set in a Dockerfile.
You can also use CUDA images, which set these variables automatically. See the
official [CUDA images](https://catalog.ngc.nvidia.com/orgs/nvidia/containers/cuda)
NGC catalog page.

87
content/manuals/engine/containers/resource_constraints.md
View File

@@ -1,8 +1,8 @@
--- ---
title: Resource constraints title: Resource constraints
weight: 30 weight: 30
description: Specify the runtime options for a container description: Limit container memory and CPU usage with runtime configuration flags
keywords: docker, daemon, configuration, runtime keywords: resource constraints, memory limits, CPU limits, cgroups, OOM, swap, docker run, memory swap
aliases: aliases:
- /engine/admin/resource_constraints/ - /engine/admin/resource_constraints/
- /config/containers/resource_constraints/ - /config/containers/resource_constraints/
@@ -265,84 +265,5 @@ If the kernel or Docker daemon isn't configured correctly, an error occurs.
## GPU ## GPU
### Access an NVIDIA GPU For information on how to access NVIDIA GPUs from a container, see
[GPU access](gpu.md).
#### Prerequisites
Visit the official [NVIDIA drivers page](https://www.nvidia.com/Download/index.aspx)
to download and install the proper drivers. Reboot your system once you have
done so.
Verify that your GPU is running and accessible.
#### Install nvidia-container-toolkit
Follow the official NVIDIA Container Toolkit [installation instructions](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html).
#### Expose GPUs for use
Include the `--gpus` flag when you start a container to access GPU resources.
Specify how many GPUs to use. For example:
```console
$ docker run -it --rm --gpus all ubuntu nvidia-smi
```
Exposes all available GPUs and returns a result akin to the following:
```bash
+-------------------------------------------------------------------------------+
| NVIDIA-SMI 384.130 Driver Version: 384.130 |
|-------------------------------+----------------------+------------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+========================|
| 0 GRID K520 Off | 00000000:00:03.0 Off | N/A |
| N/A 36C P0 39W / 125W | 0MiB / 4036MiB | 0% Default |
+-------------------------------+----------------------+------------------------+
+-------------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |
|===============================================================================|
| No running processes found |
+-------------------------------------------------------------------------------+
```
Use the `device` option to specify GPUs. For example:
```console
$ docker run -it --rm --gpus device=GPU-3a23c669-1f69-c64e-cf85-44e9b07e7a2a ubuntu nvidia-smi
```
Exposes that specific GPU.
```console
$ docker run -it --rm --gpus '"device=0,2"' ubuntu nvidia-smi
```
Exposes the first and third GPUs.
> [!NOTE]
>
> NVIDIA GPUs can only be accessed by systems running a single engine.
#### Set NVIDIA capabilities
You can set capabilities manually. For example, on Ubuntu you can run the
following:
```console
$ docker run --gpus 'all,capabilities=utility' --rm ubuntu nvidia-smi
```
This enables the `utility` driver capability which adds the `nvidia-smi` tool to
the container.
Capabilities as well as other configurations can be set in images via
environment variables. More information on valid variables can be found in the
[nvidia-container-toolkit](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/docker-specialized.html)
documentation. These variables can be set in a Dockerfile.
You can also use CUDA images, which set these variables automatically. See the
official [CUDA images](https://catalog.ngc.nvidia.com/orgs/nvidia/containers/cuda)
NGC catalog page.