
Thanks to Unraid, you can create your own media server using Jellyfin which is a free and open-source equivalent of Plex.
The advantage of Jellyfin is that video transcoding using a graphics card (GPU) does not require a (paid) license, unlike its competitors: Plex and Emby.
In this tutorial, you will see how to install and configure Jellyfin, but also how to install the prerequisites necessary for using video transcoding via a graphics card (although this is optional).
In this tutorial, we will use an "MSI GeForce RTX 3070 VENTUS 3X OC" graphics card, but the procedure is identical for all NVIDIA graphics cards supporting NVENC (NVIDIA encoder).
To check if your graphics card supports NVENC, go to the "Video Encode and Decode GPU Support Matrix" page on the official NVIDIA website.
Next, open the "NVENC - Encoding" section and select the type of graphics card:
Note that the number of concurrent sessions (or "processes" in the rest of this tutorial) is limited to 3 for "Consumer (GeForce)" graphics cards and is often unlimited for "Professional (NVIDIA RTX / Quadro)" graphics cards. .
This information is displayed in the "Max # of concurrent sessions" column of the NVIDIA table.

By default, Jellyfin will use your processor (CPU) for transcoding (converting) videos to a different quality than the original.
If you want to use hardware acceleration (hardware transcoding) using your NVIDIA graphics card (which is free with Jellyfin), you will first need to install the Unraid version of the NVIDIA driver for your graphics card(s).
To install the NVIDIA driver for your graphics card on Unraid, you will need to install the "Nvidia Driver" plugin.
To do this, search for and install the "Nvidia Driver" plugin via "Community Applications".
Note that the "Unraid Nvidia" plugin created by "linuxserver" no longer exists. Nonetheless, the "Nvidia Driver" plugin created by "ich777" is a very similar replacement to "Unraid Nvidia".

Important : downloading and installing this "Nvidia Driver" plugin will also download the Nvidia driver.
Which may take a few minutes depending on your Internet connection.
However, this will only happen during the first installation of the plugin and not during updates of this "Nvidia Driver" plugin.
It is therefore important to leave this window open until the "Done" button (orange) appears.

Once the Nvidia driver is downloaded and installed, the "Done" button will appear.
Now click on this "Done" button.
Note that the author of this plugin tells you that you will need to disable and then re-enable Docker if you have just installed the NVIDIA driver for the first time.

After installing the "Nvidia Driver" plugin, a red notification will be displayed to warn you again that you need to restart Docker following the installation of the NVIDIA driver.
To do this, go to: Settings.

Next, click on Docker.

Change the value of the "Enable Docker" parameter to "No".

Then, click "Apply".

As you can see on the right, Docker is now stopped (status: stopped).
Re-enable Docker by changing "Enable Docker" to "Yes".

Then, click: Apply.

Docker is running again (status: running).

To access information about the NVIDIA driver installed on Unraid, go to: Settings -> Nvidia Driver.

On the "Nvidia Driver Package" page that appears, you will find:
On the right, you will have the possibility to choose another version of NVIDIA driver if you wish. However, we advise you to only use production versions to avoid system instabilities.

To ensure that Jellyfin is using your graphics card for video transcoding and not your processor, you can install the "GPU Statistics" plugin.
With this, you will see the overall usage percentage of your graphics card, as well as some of its features (including hardware encoding/decoding).
Search for and install the "GPU Statistics" plugin from "b3rs3rk" via "Community Applications".

Wait while this "GPU Statistics" plugin downloads and installs.

To configure the "GPU Statistics" plugin and in particular choose which graphics card you want to monitor, go to: Settings -> GPU Statistics.

The "GPU Statistics" page appears.

To get started, select "Vendor: NVIDIA".

Next, from the "Unit ID for Dashboard" list, select the graphics card you will use for hardware transcoding via Jellyfin.
In our case, our "NVIDIA GeForce RTX 3070" graphics card.
Note that the "GPU-xxxxxxxx-xxxxxx-xxxxxxxx" identifier behind the name of your graphics card corresponds to its UUID.
This UUID will be needed when configuring the "Jellyfin" container if you want to use hardware transcoding.
But, we'll come back to that later.

Then click: Apply.

The "GPU Statistics" plugin appears in the list of plugins.

NAS 1/15/2026
NAS 2/11/2026
NAS 1/1/2026
NAS 4/1/2026
Pinned content
Contact
® InformatiWeb-Pro.net - InformatiWeb.net 2008-2022 - © Lionel Eppe - All rights reserved.
Total or partial reproduction of this site is prohibited and constitutes an infringement punishable by articles L.335-2 and following of the intellectual property Code.
You must be logged in to post a comment