Having problems with TVHeadend/vaapi hardware encoding.
I was running Unraid 6.12.5 with Intel i5-9600t (UHD 630 iGPU); In preparation for hardware upgrade to i7-11700t (UHD 750 iGPU) I updated unRAID to 7.2.3 ( incremental 7.0, 7.1. 7.2, 7.2.3), and updated all plugins including ich777's Intel GPU Top. Before this software update, TVHeadEnd was recording in h265 vaapi just fine, but the the Intel_GPU_Top I was running was probably couple years old.

Unfortunately, and this is my fault, I didn't check (I did so many other things) if just the software components update broke this or not.

Upgraded the hardware and it wasn't till late in the night to discover that TVHeadend records the audio but video is not usable by any player, pic:


Surprisingly, Jellyfin transcodes with iGPU just fine. However, I did notice some differences - TVHeadEnd uses the "VideoEnhance" while the Jellyfin does not:
Jellyfin (works):


TVHeadEnd (no longer works):


It is worth mentioning that the originally I stayed with the same TVHeadEnd docker container. When I first discovered the issue, in the TVHeadEnd docker log, I saw a warning that the read something of the sort "No VA Display, unable to find /dev/dri/renderD128. So googled a bit and executed this permissions command:
chmod 777 /dev/dri/*

This solved the no renderD128 error in TVHE log. However, the result are the same. I went down the rabbit hole searching for problems within TVHeadEnd - numerous reports of failed vaapi encoding. Spent a day on that, udapted to latest and dowgraded to older builds - no changes.
Does this seem right for dev/dri list inside the container:
root:~# sudo docker exec -it tvheadend ls -l /dev/dri

total 0

crwxrwxrwx 1 root audio 226, 0 Jan 20 19:08 card0

crwxrwxrwx 1 root audio 226, 128 Jan 20 19:08 renderD128

Anyone has experience this or can at least point me to what should I do next?

Edited January 21Jan 21 by OrdinaryButt

5 hours ago, OrdinaryButt said:

So googled a bit and executed this permissions command:
chmod 777 /dev/dri/*

This is not necessary on Unraid.

5 hours ago, OrdinaryButt said:

I saw a warning that the read something of the sort "No VA Display, unable to find /dev/dri/renderD128

This indicates that something is wrong with your transcoding settings or the passthrough to the container.

5 hours ago, OrdinaryButt said:

ich777's Intel GPU Top. Before this software update, TVHeadEnd was recording in h265 vaapi just fine, but the the Intel_GPU_Top I was running was probably couple years old.

You have to understand that intel_gpu_top is just a binary which can display the utilization from your iGPU/dGPU and has nothing to do with transcoding nor the driver itself. The drivers ship with the Kernel and therefore Unraid.

It's most likely that something in terms of the system upgrade causes that since as said, the plugin is just a binary to display the utilization.

Since this seems not related to the Intel-GPU-TOP plugin, please create a post in the TVHeadend Container support thread.

9 hours ago, ich777 said:

This is not necessary on Unraid.

  14 hours ago, OrdinaryButt said:

I saw a warning that the read something of the sort "No VA Display, unable to find /dev/dri/renderD128

This indicates that something is wrong with your transcoding settings or the passthrough to the container.

It is not necessary but when I changed the permissions with chmod 777 /dev/dri/* I stopped getting the "No VA display/unable to find /dev/dri/renderD128 in TVHE docker log.

Worth mentioning that original unRAID/TVHE setup was from ~2021. Wasn't there some plugin/docker permission handling change at some point?

4 hours ago, OrdinaryButt said:

Worth mentioning that original unRAID/TVHE setup was from ~2021. Wasn't there some plugin/docker permission handling change at some point?

Unraid is handling the permissions on boot and Intel GPU TOP too.

On 1/12/2026 at 3:46 AM, Jclendineng said:

Just so I’m clear, this isn’t intended for GPU stats, mostly iGpu. My a380 worked fine with it but my b50 pro stats aren’t showing up so I should be good to uninstall this?

Intel GPU top is for i915 driver Bxx uses the xe driver and not covered by top, there is talk about using for xe but no timescales.

I got AOOSTAR WTR MAX with AMD R7 PRO 8845HS CPU + AMD Radeon 780M.

With radeontop I get:

root@Tower:~# radeontop -d -

Unknown Radeon card. <= R500 won't work, new cards might.

Dumping to -, until termination.

1770115276.772174: bus 01, gpu 0.00%, ee 0.00%, vgt 0.00%, ta 0.00%, tc 0.00%, sx 0.00%, sh 0.00%, spi 0.00%, smx 0.00%, cr 0.00%, sc 0.00%, pa 0.00%, db 0.00%, cb 0.00%, vram 97.77% 425.46mb, gtt 1.67% 261.64mb, mclk 100.00% 2.800ghz, sclk 29.63% 0.800ghz

1770115277.772290: bus 01, gpu 0.00%, ee 0.00%, vgt 0.00%, ta 0.00%, tc 0.00%, sx 0.00%, sh 0.00%, spi 0.00%, smx 0.00%, cr 0.00%, sc 0.00%, pa 0.00%, db 0.00%, cb 0.00%, vram 97.77% 425.46mb, gtt 1.67% 261.64mb, mclk 100.00% 2.800ghz, sclk 29.63% 0.800ghz

1770115278.772392: bus 01, gpu 0.00%, ee 0.00%, vgt 0.00%, ta 0.00%, tc 0.00%, sx 0.00%, sh 0.00%, spi 0.00%, smx 0.00%, cr 0.00%, sc 0.00%, pa 0.00%, db 0.00%, cb 0.00%, vram 97.77% 425.45mb, gtt 1.67% 261.64mb, mclk 100.00% 2.800ghz, sclk 29.63% 0.800ghz

1770115279.772486: bus 01, gpu 0.00%, ee 0.00%, vgt 0.00%, ta 0.00%, tc 0.00%, sx 0.00%, sh 0.00%, spi 0.00%, smx 0.00%, cr 0.00%, sc 0.00%, pa 0.00%, db 0.00%, cb 0.00%, vram 97.77% 425.46mb, gtt 1.67% 261.64mb, mclk 100.00% 2.800ghz, sclk 29.63% 0.800ghz

1770115280.214893: bus 01, gpu 0.00%, ee 0.00%, vgt 0.00%, ta 0.00%, tc 0.00%, sx 0.00%, sh 0.00%, spi 0.00%, smx 0.00%, cr 0.00%, sc 0.00%, pa 0.00%, db 0.00%, cb 0.00%, vram 97.77% 425.45mb, gtt 1.67% 261.64mb, mclk 100.00% 2.800ghz, sclk 29.63% 0.800ghz

No actual usage shown. This is happening while I am getting 500 FPS in Tdarr while CPU is <5%.

In GPU Statistics I see and this plugin is also showing 0% GPU Load:

Is this CPU/GPU not compatible?

Edit: I found some actual usage output with:

root@Tower:~# cat /sys/kernel/debug/dri/0/amdgpu_pm_info

GFX Clocks and Power:

2800 MHz (MCLK)

800 MHz (SCLK)

0 MHz (PSTATE_SCLK)

0 MHz (PSTATE_MCLK)

974 mV (VDDGFX)

965 mV (VDDNB)

19.180 W (average SoC including CPU)

24.43 W (current SoC including CPU)

GPU Temperature: 50 C

GPU Load: 0 %

VCN Load: 99 %

SMC Feature Mask: 0x07a5f2f3f3fff7ff

VCN: Powered up

Clock Gating Flags Mask: 0x3b8099301

Graphics Fine Grain Clock Gating: On

Graphics Medium Grain Clock Gating: On

Graphics Medium Grain memory Light Sleep: Off

Graphics Coarse Grain Clock Gating: Off

Graphics Coarse Grain memory Light Sleep: Off

Graphics Coarse Grain Tree Shader Clock Gating: Off

Graphics Coarse Grain Tree Shader Light Sleep: Off

Graphics Command Processor Light Sleep: Off

Graphics Run List Controller Light Sleep: Off

Graphics 3D Coarse Grain Clock Gating: Off

Graphics 3D Coarse Grain memory Light Sleep: Off

Memory Controller Light Sleep: On

Memory Controller Medium Grain Clock Gating: On

System Direct Memory Access Light Sleep: Off

System Direct Memory Access Medium Grain Clock Gating: Off

Bus Interface Medium Grain Clock Gating: On

Bus Interface Light Sleep: On

Unified Video Decoder Medium Grain Clock Gating: Off

Video Compression Engine Medium Grain Clock Gating: Off

Host Data Path Light Sleep: On

Host Data Path Medium Grain Clock Gating: On

Digital Right Management Medium Grain Clock Gating: Off

Digital Right Management Light Sleep: Off

Rom Medium Grain Clock Gating: Off

Data Fabric Medium Grain Clock Gating: Off

VCN Medium Grain Clock Gating: Off

Host Data Path Deep Sleep: Off

Host Data Path Shutdown: Off

Interrupt Handler Clock Gating: On

JPEG Medium Grain Clock Gating: Off

Repeater Fine Grain Clock Gating: On

Perfmon Clock Gating: On

Address Translation Hub Medium Grain Clock Gating: On

Address Translation Hub Light Sleep: On

Edited February 3Feb 3 by mrbusiness

I'm updating from 7.1.4 to 7.2.3 and I received the following 2 messages from Plugin Update Helper

Coral-Driver download failed, please go to the support thread for this plugin and make a post with a screenshot from this error!

Download from plugin package(s): Coral-Driver for unRAID v7.2.3 failed! Please visit the support thread(s) before rebooting to avoid plugin issues!

I've been unable to find any additional information in Unraid as to what the actual error is and how I might resolve it prior to rebooting.

Edited February 8Feb 8 by NSX

1 hour ago, NSX said:

I've been unable to find any additional information in Unraid as to what the actual error is and how I might resolve it prior to rebooting.

Please try to reboot, the plugin should try to download the driver on boot too.
However the driver for 7.2.3 should be available, maybe there is some issue on your network, please check if you are maybe blocking access to the GitHub API.

If the plugin doesn't work after the reboot, please try to uninstall the plugin, rebooting and installing the plugin again.

9 hours ago, ich777 said:

Please try to reboot, the plugin should try to download the driver on boot too.
However the driver for 7.2.3 should be available, maybe there is some issue on your network, please check if you are maybe blocking access to the GitHub API.

If the plugin doesn't work after the reboot, please try to uninstall the plugin, rebooting and installing the plugin again.

I think things resolved on a reboot. There wouldn't be anything on my network causing issues or blocking Internet access.

I saw this on reboot and Frigate seems to be able to detect the TPU w/o an issue

3 hours ago, NSX said:

I saw this on reboot and Frigate seems to be able to detect the TPU w/o an issue

Yes, this was the download on boot, maybe you had no GitHub Tokens left or some kind of network issue occurred when doing the update.

Glad to hear that it works. :)

My question is regarding the plugin UGREEN LED Driver:

I have a UGREEN DXP2800 with 3 LEDs.

After installing Unraid and UGREEN LED Driver, I face following behavior:

LED1 (left): Going on/off periodically, lightning sometimes short, sometimes long

LED2 (middle): Always on, even if HDD1 is spinned down

LED3 (right): Always on, even if HDD2 is spinned down

My question: How should the LEDs behave and what does it show?

31 minutes ago, mafa1410 said:

LED1 (left): Going on/off periodically, lightning sometimes short, sometimes long

First LED is LAN activity and will blink, or light up if activity is detected.

32 minutes ago, mafa1410 said:

LED2 (middle): Always on, even if HDD1 is spinned down

LED3 (right): Always on, even if HDD2 is spinned down

Spin down won't disable LEDs on Unraid for disks.

Does the intel_gpu_top plug-in install the Intel Compute Runtime ( intel/compute-runtime: Intel® Graphics Compute Runtime for oneAPI Level Zero and OpenCL™ Driverintel/compute-runtime: Intel® Graphics Compute Runtime for oneAPI Level Zero and OpenCL™ Driver )? When I look in /etc/OpenCL/vendors, I don't see intel.icd listed and it doesn't show up when clinfo (installed with un-get) is run in the host. Is there any change this runtime could be included as well?

6 hours ago, millercentral said:

Does the intel_gpu_top plug-in install the Intel Compute Runtime

No, why should it?

6 hours ago, millercentral said:

Is there any change this runtime could be included as well?

No, the plugin is only meant for Intel GPU TOP

However why would you need the runtime on the host? Usually Docker containers or whatever application comes with the necessary runtimes.

I was having a problem getting OpenCL to work on my intel GPU in a container and assumed it was because this was not installed in the host (it is installed in the host for the Nvidia GPU in the same system). However, the problem turned out to be that if you pass both an Nvidia GPU (via the --runtime nvidia) AND the Intel GPU (via --device /dev/dri:/dev/dri) then the OpenCL driver will ONLY work for the Nvidia GPU and not the Intel one. If you only pass the Intel GPU and not the Nvidia one, then the Intel OpenCL driver in the container works fine. I confirmed this with several different containers.

30 minutes ago, millercentral said:

I was having a problem getting OpenCL to work on my intel GPU in a container and assumed it was because this was not installed in the host (it is installed in the host for the Nvidia GPU in the same system).

For Intel that is not the case, the container must ship the libraries/runtimes and you are good to go.

31 minutes ago, millercentral said:

However, the problem turned out to be that if you pass both an Nvidia GPU (via the --runtime nvidia) AND the Intel GPU (via --device /dev/dri:/dev/dri) then the OpenCL driver will ONLY work for the Nvidia GPU and not the Intel one. If you only pass the Intel GPU and not the Nvidia one, then the Intel OpenCL driver in the container works fine. I confirmed this with several different containers.

I would recommend that you report that in the support threads from the containers, maybe there is a workaround for this since you should be able to do use one or the other or even both GPUs depending about which container we are talking about. However as said this is something for the support threads from the container(s).

I feel like this might be super simple but I'm having trouble figuring it out.

After installing the Radeon Top CA plugin when I run the command I get 0% for everything:

The GPU Diagnostic data in the dashboard shows the same:

Even though it seems to be detecting the card:

I am wondering if I need to do anything special with my BIOS as it groups all my PCI devices in TOOLS > System Devices. I do see the card under PCI Devices - (No IOMMU Groups Available).

I also attached my diagnostics.

tower-diagnostics-20260311-1330.zip

2 hours ago, deemuns said:

After installing the Radeon Top CA plugin when I run the command I get 0% for everything:

If nothing is using the card nothing will show up and you will see 0%

So in your case nothing is using the card.

14 hours ago, ich777 said:

If nothing is using the card nothing will show up and you will see 0%

So in your case nothing is using the card.

Thank you for the quick response. I was thinking there was a problem because the GPU Statistics dashboard widget was not showing a vendor, driver, or device.

I was also able to passthrough the GPU in my plex container and select it within the plex transcoding menu but it does not appear to be doing anything when I force plex to transcode something that is playing. It does not have the (hw) next to the item to show it is being hardware transcoded.

38 minutes ago, deemuns said:

I was also able to passthrough the GPU in my plex container and select it within the plex transcoding menu but it does not appear to be doing anything when I force plex to transcode something that is playing. It does not have the (hw) next to the item to show it is being hardware transcoded.

I would recommend that you post about that issue in the appropriate support thread.

FWIK Plex isn't the best choice when it comes to HW transcoding on AMD hardware.

However radeontop will show you when something is using your GPU.

Hello

I was hoping to get help with the Unraid QNAP-EC plugin. I'm using a tvs-h1688x and I installed the QNAP-EC but it says it's not available. Any help would be appreciated.

1 hour ago, Pabloicecreambar said:

I was hoping to get help with the Unraid QNAP-EC plugin. I'm using a tvs-h1688x and I installed the QNAP-EC but it says it's not available. Any help would be appreciated.

You can only create an issue here and request support for this device:
https://github.com/Stonyx/QNAP-EC

When will the intel gpu top support battlemage cards?

20 minutes ago, T_Matz said:

When will the intel gpu top support battlemage cards?

Most likely never. They do refer to use other tools and the issues are closed.

16 minutes ago, Mainfrezzer said:

Most likely never. They do refer to use other tools and the issues are closed.

Oh I wasnt tracking that! I know my battlemage card works i just cant see any details on whats happeneing with it. Ill try and find these other tools you are talkingabout unless you have more details.