kemko/nomad
1
0
Fork 0
mirror of https://github.com/kemko/nomad.git synced 2026-01-06 10:25:42 +03:00
Code Issues Packages Projects Releases Wiki Activity

nvidia driver: add MIG support to overview paragraph (#24099)

Browse Source
This commit is contained in:
Aimee Ukasick
2024-10-03 09:08:43 -05:00
committed by GitHub
parent 1fabbaa179
commit e5b18affa1
1 changed files with 27 additions and 70 deletions
Download Patch File Download Diff File Expand all files Collapse all files

97
website/content/plugins/devices/nvidia.mdx
View File

@@ -8,80 +8,34 @@ description: The Nvidia Device Plugin detects and makes Nvidia devices available
Name: `nomad-device-nvidia`
The Nvidia device plugin is used to expose Nvidia GPUs to Nomad.
Use the NVIDIA device plugin to expose NVIDIA GPUs to Nomad. The driver
automatically supports [Multi-Instance GPU (MIG)][mig].
The NVIDIA device plugin uses [NVML] bindings to get data regarding available
NVIDIA devices and then exposes them via [Fingerprint RPC]. The plugin detects
whether the GPU has Multi-Instance GPU enabled, and when enabled, the plugin
fingerprints all instances as individual GPUs. You may exclude GPUs from
fingerprinting by setting the [`ignored_gpu_ids` field](#plugin-configuration).
## Fingerprinted Attributes
<table>
<thead>
<tr>
<th>Attribute</th>
<th>Unit</th>
</tr>
</thead>
<tbody>
<tr>
<td>
<tt>memory</tt>
</td>
<td>MiB</td>
</tr>
<tr>
<td>
<tt>power</tt>
</td>
<td>W (Watt)</td>
</tr>
<tr>
<td>
<tt>bar1</tt>
</td>
<td>MiB</td>
</tr>
<tr>
<td>
<tt>driver_version</tt>
</td>
<td>string</td>
</tr>
<tr>
<td>
<tt>cores_clock</tt>
</td>
<td>MHz</td>
</tr>
<tr>
<td>
<tt>memory_clock</tt>
</td>
<td>MHz</td>
</tr>
<tr>
<td>
<tt>pci_bandwidth</tt>
</td>
<td>MB/s</td>
</tr>
<tr>
<td>
<tt>display_state</tt>
</td>
<td>string</td>
</tr>
<tr>
<td>
<tt>persistence_mode</tt>
</td>
<td>string</td>
</tr>
</tbody>
</table>
| Attribute | Unit |
|------------------|----------|
| memory | MiB |
| power | W (Watt) |
| bar1 | MiB |
| driver_version | string |
| cores_clock | MHz |
| memory_clock | MHz |
| pci_bandwidth | MB/s |
| display_state | string |
| persistence_mode | string |
## Runtime Environment
The `nvidia-gpu` device plugin exposes the following environment variables:
- `NVIDIA_VISIBLE_DEVICES` - List of Nvidia GPU IDs available to the task.
- `NVIDIA_VISIBLE_DEVICES` - List of NVIDIA GPU IDs available to the task.
### Additional Task Configurations
@@ -101,7 +55,7 @@ In order to use the `nomad-device-nvidia` device driver the following prerequisi
### Container Toolkit Installation
Follow the [NVIDIA Container Toolkit installation instructions][nvidia_container_toolkit]
from Nvidia to prepare a machine to use docker containers with Nvidia GPUs. You should
from NVIDIA to prepare a machine to use docker containers with NVIDIA GPUs. You should
be able to run this simple command to test your environment and produce meaningful
output.
@@ -135,8 +89,8 @@ config:
## Limitations
The Nvidia integration only works with drivers who natively integrate with
Nvidia's [container runtime
The NVIDIA integration only works with drivers who natively integrate with
NVIDIA's [container runtime
library](https://github.com/NVIDIA/libnvidia-container).
Nomad has tested support with the [`docker` driver][docker-driver].
@@ -273,7 +227,10 @@ Wed Jan 23 18:25:32 2019
+-----------------------------------------------------------------------------+
```
[NVML]: https://github.com/NVIDIA/go-nvml
[Fingerprint RPC]:
/nomad/docs/concepts/plugins/devices#fingerprint-context-context-chan-fingerprintresponse-error
[mig]: https://www.nvidia.com/en-us/technologies/multi-instance-gpu/
[docker-driver]: /nomad/docs/drivers/docker 'Nomad docker Driver'
[exec-driver]: /nomad/docs/drivers/exec 'Nomad exec Driver'
[java-driver]: /nomad/docs/drivers/java 'Nomad java Driver'