N150 iGPU Passthrough

HA99

New Member
Mar 16, 2025
2
0
1
I am brand new to Proxmox and thought I would give it a try when I saw a video that setup a Jellyfin container on a mini PC with an Intel N100 CPU. I want to do almost the exact same thing, except that I have the newer N150 GPU. I've spent the past two days reading numerous blogs, tutorials and watching videos but I can't get it to work.

I've enabled the no-subscription repositories, and read in one of the blogs that I needed to update the kernel to 6.11.11, which I have done as well.

The main tutorial I have been following is this one from WunderTech: https://www.wundertech.net/installing-jellyfin-on-proxmox/. I've added the suggested intel drivers on the host with this:
Code:
apt install software-properties-common -y
add-apt-repository -y non-free
apt install intel-media-va-driver-non-free -y

When I run vainfo, none of the VAProfiles are shown; all I get is this:
Code:
root@pve:~# vainfo
Trying display: wayland
Trying display: x11
error: can't connect to X server!
Trying display: drm
libva info: VA-API version 1.20.0
libva info: Trying to open /usr/lib/x86_64-linux-gnu/dri/iHD_drv_video.so
libva info: Found init function __vaDriverInit_1_17
libva error: /usr/lib/x86_64-linux-gnu/dri/iHD_drv_video.so init failed
libva info: va_openDriver() returns 1
libva info: Trying to open /usr/lib/x86_64-linux-gnu/dri/i965_drv_video.so
libva info: va_openDriver() returns -1
vaInitialize failed with error code -1 (unknown libva error),exit

I haven't been able to find a resolution for the error. Any ideas?
 
I feel your pain , I've got a new n150 exactly the same for me . I followed the same tutorial and probably some of the others you have tried. In the video he makes a special note to explain specifically that he is doing this with the n100 and that other processors may not work, as we have found .

I'm an absolute Linux numpty, but I have a feeling the kernel just ain't there yet.

I tried direct arch , without proxmox and installed Jellyfin and the intel drivers with some success . GPU was confirmed as working through control centre or similar app that showed the GPU graph .

However I could only get it working on the N150 box , I had zero luck in getting Jellyfin to transcode out to other machines.

I think the n150 is too new and we may just have to wait until kernel support catches up .

If anyone gets it working please share , there must be others out there experiencing the same issue.

A problem shared is a problem solved
 
Last edited:
Hi, I am in the same boat as you are ... however, I somehow managed to get the passthrough. But I fail to connect the passthrough to the physically connected HDMI (to e.g. get direct output to my TV).
x-igd-opregion=on
I get:
IGD OpRegion not supported
I looked through the BIOS and just do not quite understand if this feature can be turned on or not?
 
Hi, I am in the same boat as you are ... however, I somehow managed to get the passthrough. But I fail to connect the passthrough to the physically connected HDMI (to e.g. get direct output to my TV).
x-igd-opregion=on
I get:
IGD OpRegion not supported
I looked through the BIOS and just do not quite understand if this feature can be turned on or not?
When you pass through is work , how are you verifying this ? If you're running Jellyfin , did you get transcoding working to other player devices while from your n150 ?
 
Hi, I am in the same boat as you are ... however, I somehow managed to get the passthrough. But I fail to connect the passthrough to the physically connected HDMI (to e.g. get direct output to my TV).
x-igd-opregion=on
I get:
IGD OpRegion not supported
I looked through the BIOS and just do not quite understand if this feature can be turned on or not?
I saw your reply in another thread today with details on what you did; thanks for that. I didn't want to hijack that thread (I'm not looking to passthrough the HDMI port, nor install a Win 11 VM), but the first two steps *might* be what I need.

I have some doubts on the instructions you linked to where he references kernel 6.8 and I've read elsewhere that 6.11 is needed, as well as his warnings about updates requiring the process to be redone (I'm a linux novice). He also mentions Intel driver issues with LXC, which I don't understand since there are many online references showing this works.

At any rate, I'll give it a try when I'm home later this week and see if it works.
 
For what it's worth, I'm in exactly the same boat - Jellyfin works fine until I start transcoding and then... it just stops. And now the host /dev/dri/renderD128 file has just disappeared (this is what comes of following instructions for an N100!). Time to re-install Proxmox.
 
100%

I'm thinking it's probably best to cough up the cash for a n100 and see how that goes.

N150 can go on the shelf for a while until Linux catches up.

My fault, i shouldn't have expected it to just work so easily, new hardware needs time.
 
Honestly, I like the N150 very much and as a bare metal it works pretty decent. On one instance I even got the Ollama ipex running with a 7bn model :) Speed is ok, but sure there is upside potential with an other GPU.
Using the 6.11.11 kernel from the repository isn't that big of a deal - compiling the i915 module is pretty straight forward (copy paste).

HDMI problem:
I tried to turn it intu a LDME6 Workstation (with Proxmox) but here once more is the problem remains HDMI output and passthrough :( ... I also seem not to be able to properly set up the intel drivers. Xorg always falls back to software rendering. Vainfo stops with segmentation faults on the main display.
Somwhere I heard that when trying to install Ubuntu on the N150 you need kernel 6.12 to get it properly working ... well, with Proxmox this is probably hell to compile yourself.

hdmi-output-from-vm-with-igpu-passthrough-n150-alder-lake-v0-sp4c83w4phoe1.webp
hdmi-output-from-vm-with-igpu-passthrough-n150-alder-lake-v0-c8n4c9n2phoe1.webp