ZimaBoard2 with Jellyfin - High CPU utilisation

[sehHeiden]

Hi,

finally got my ZimaBoard2 with rack last week. This week I got my hdds. So I wanted to start. One of the fist apps I installed is Jellyfin.

For few days now. I uses a single core for ffmpeg tasks on a single core. Gets relatively hot 80 to 85 degrees in btop, but no gpu utilisation in btop shown.
Do I need to turn the GPU use on somewhere? Like in the docker settings (E.g. change the tag used?)
I hope, that with GPU usage it would be a) faster and b) more energy efficient (perhaps even cooler?)

But looks like I need to disassemble everything the next days to actually add the fan. At least I hope it helps, against these temperatures. Does anyone have experience how much that cools and how loud it is?

[Zima-Giorgio]

I think this post may be helpful to your question.

Jellyfin HDR Tone mapping Hardware Acceleration

Regarding Jellyfin, the current version available in the app store seems to have some issues with color tone conversion. I recommend using the version from https://hub.docker.com/r/nyanmisaka/jellyfin You can use my configuration file for a customized installation: Jellyfin.yaml (8.8 KB) Here are my transcoding settings: [image] [image] This is the final playback result: [image] Since the integrated graphics performance of the N100 is not very strong, it seems to only handle smooth co…

The key point is that you need to install a version of Jellyfin that is capable of hardware acceleration/transcoding.

Look forward to your feedback.

[sehHeiden]

@Jellyfin HDR Tone mapping Hardware Acceleration

are my transcoding setting

So the old installation allowed access to the drivers, then I activated IntelQuickSense for Transcoding, and HW for Trickplay.

I acctivated Tone Mapping, BUT almost all my data is in 720p or lower. Hence, I assume it’s not important. I did not activate Low Power Encoding, as I do not know whether this will work?

@Zima-Giorgio I assume, the problem you mentioned is not important for my usecase?

Still, I notice a temperature reduction by 30 degrees to 51 degree Celsius.
Currently, I am wondering why the drives never stop? I do not see that any task is using the drive. For jellyfin there is still a ffmpeg task, but that is at 0.0 CPU utilization.