AMD Mi50 : unable to add GPU to Ollama app

janot928 · September 20, 2025, 12:27am

Hello everyone,

I’ve recently added a new GPU to my truenas server (AMD Mi50 32Gb) to replace my good old Nvidia P2000

The problem i’m having is that i cannot add it to the Ollama App : i get this error message (using both the ROCM image and the default image)

Traceback (most recent call last):
File “/usr/lib/python3/dist-packages/middlewared/job.py”, line 515, in run
await self.future
File “/usr/lib/python3/dist-packages/middlewared/job.py”, line 562, in __run_body
rv = await self.middleware.run_in_thread(self.method, *args)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File “/usr/lib/python3/dist-packages/middlewared/main.py”, line 627, in run_in_thread
return await self.run_in_executor(io_thread_pool_executor, method, *args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File “/usr/lib/python3/dist-packages/middlewared/main.py”, line 624, in run_in_executor
return await loop.run_in_executor(pool, functools.partial(method, *args, **kwargs))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File “/usr/lib/python3.11/concurrent/futures/thread.py”, line 58, in run
result = self.fn(*self.args, **self.kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File “/usr/lib/python3/dist-packages/middlewared/service/crud_service.py”, line 294, in nf
rv = func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File “/usr/lib/python3/dist-packages/middlewared/api/base/decorator.py”, line 101, in wrapped
result = func(*args)
^^^^^^^^^^^
File “/usr/lib/python3/dist-packages/middlewared/plugins/apps/crud.py”, line 229, in do_update
app = self.update_internal(job, app, data, trigger_compose=app[‘state’] != ‘STOPPED’)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File “/usr/lib/python3/dist-packages/middlewared/plugins/apps/crud.py”, line 267, in update_internal
compose_action(app_name, app[‘version’], ‘up’, force_recreate=True, remove_orphans=True)
File “/usr/lib/python3/dist-packages/middlewared/plugins/apps/compose_utils.py”, line 61, in compose_action
raise CallError(err_msg)
middlewared.service_exception.CallError: [EFAULT] Failed ‘up’ action for ‘ollama’ app. Please check /var/log/app_lifecycle.log for more details

When checking the afformentionned log file, I have this line

[2025/09/20 02:04:05] (ERROR) app_lifecycle.compose_action():56 - Failed ‘up’ action for ‘ollama’ app: Network ix-ollama_default Creating\n Network ix-ollama_default Created\n Container ix-ollama-ollama-1 Creating\n Container ix-ollama-ollama-1 Created\n Container ix-ollama-ollama-1 Starting\nError response from daemon: error gathering device information while adding custom device “/dev/dri”: no such file or directory\n

Indeed, trying to do a “cd /dev/dri” doesn’t lead anywhere and there isn’t a dri folder in the dev folder

Looking quickly around i’ve found a post talking about Iommu beeing enable for the GPU beeing the cause of this no “/dev/dri” error.
https://forums.unraid.net/topic/148092-steps-to-get-plex-hardware-transcoding-to-work-with-amd-igpu-vega-on-amd-mini-pc/

I had some trouble passing my previous GPU (Nvidia P2000) to the apps, but got it working following this post : https://forums.truenas.com/t/docker-apps-and-uuid-issue-with-nvidia-gpu-after-upgrade-to-24-10-or-25-04/22547 . Could it be something similar ?

Before going any further, I would really like some feedback before breacking anything

Thanks in advance !

Some info on the server
Cpu : AMD Threadripper W2290X
Ram : a mish mash of kits running on stock DDR4 speed ( 2 x 2 x 32gb + 4x8 gb) for a total of 157.3 Gib of usable Ram
MoBo : Asus X399-e Gaming
1 pci x16 : AMD Mi50 32gb
2 pci x16 : empty
3 pci x16 : LSI 9300-16i 12Gbps SAS-3
4 pci x16 : Mellanox ConnectX-2 10Gbe

admin@truenas ~ % lspci
00:00.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Root Complex
00:00.2 IOMMU: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) I/O Memory Management Unit
00:01.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-1fh) PCIe Dummy Host Bridge
00:01.1 PCI bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) PCIe GPP Bridge
00:01.2 PCI bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) PCIe GPP Bridge
00:01.3 PCI bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) PCIe GPP Bridge
00:02.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-1fh) PCIe Dummy Host Bridge
00:03.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-1fh) PCIe Dummy Host Bridge
00:03.1 PCI bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) PCIe GPP Bridge
00:04.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-1fh) PCIe Dummy Host Bridge
00:07.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-1fh) PCIe Dummy Host Bridge
00:07.1 PCI bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Internal PCIe GPP Bridge 0 to Bus B
00:08.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-1fh) PCIe Dummy Host Bridge
00:08.1 PCI bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Internal PCIe GPP Bridge 0 to Bus B
00:14.0 SMBus: Advanced Micro Devices, Inc. [AMD] FCH SMBus Controller (rev 59)
00:14.3 ISA bridge: Advanced Micro Devices, Inc. [AMD] FCH LPC Bridge (rev 51)
00:18.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 0
00:18.1 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 1
00:18.2 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 2
00:18.3 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 3
00:18.4 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 4
00:18.5 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 5
00:18.6 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 6
00:18.7 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 7
00:19.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 0
00:19.1 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 1
00:19.2 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 2
00:19.3 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 3
00:19.4 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 4
00:19.5 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 5
00:19.6 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 6
00:19.7 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Data Fabric: Device 18h; Function 7
01:00.0 USB controller: Advanced Micro Devices, Inc. [AMD] X399 Series Chipset USB 3.1 xHCI Controller (rev 02)
01:00.1 SATA controller: Advanced Micro Devices, Inc. [AMD] X399 Series Chipset SATA Controller (rev 02)
01:00.2 PCI bridge: Advanced Micro Devices, Inc. [AMD] X399 Series Chipset PCIe Bridge (rev 02)
02:02.0 PCI bridge: Advanced Micro Devices, Inc. [AMD] 300 Series Chipset PCIe Port (rev 02)
02:03.0 PCI bridge: Advanced Micro Devices, Inc. [AMD] 300 Series Chipset PCIe Port (rev 02)
02:04.0 PCI bridge: Advanced Micro Devices, Inc. [AMD] 300 Series Chipset PCIe Port (rev 02)
02:09.0 PCI bridge: Advanced Micro Devices, Inc. [AMD] 300 Series Chipset PCIe Port (rev 02)
03:00.0 Ethernet controller: Intel Corporation I211 Gigabit Network Connection (rev 03)
06:00.0 USB controller: ASMedia Technology Inc. ASM2142/ASM3142 USB 3.1 Host Controller
07:00.0 Non-Volatile memory controller: SK hynix BC501 NVMe Solid State Drive
08:00.0 Ethernet controller: Mellanox Technologies MT26448 [ConnectX EN 10GigE, PCIe 2.0 5GT/s] (rev b0)
09:00.0 PCI bridge: PLX Technology, Inc. PEX 8724 24-Lane, 6-Port PCI Express Gen 3 (8 GT/s) Switch, 19 x 19mm FCBGA (rev ca)
0a:00.0 PCI bridge: PLX Technology, Inc. PEX 8724 24-Lane, 6-Port PCI Express Gen 3 (8 GT/s) Switch, 19 x 19mm FCBGA (rev ca)
0a:08.0 PCI bridge: PLX Technology, Inc. PEX 8724 24-Lane, 6-Port PCI Express Gen 3 (8 GT/s) Switch, 19 x 19mm FCBGA (rev ca)
0a:09.0 PCI bridge: PLX Technology, Inc. PEX 8724 24-Lane, 6-Port PCI Express Gen 3 (8 GT/s) Switch, 19 x 19mm FCBGA (rev ca)
0b:00.0 Serial Attached SCSI controller: Broadcom / LSI SAS3008 PCI-Express Fusion-MPT SAS-3 (rev 02)
0d:00.0 Serial Attached SCSI controller: Broadcom / LSI SAS3008 PCI-Express Fusion-MPT SAS-3 (rev 02)
0e:00.0 Non-Essential Instrumentation [1300]: Advanced Micro Devices, Inc. [AMD] Zeppelin/Raven/Raven2 PCIe Dummy Function
0e:00.2 Encryption controller: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Platform Security Processor (PSP) 3.0 Device
0e:00.3 USB controller: Advanced Micro Devices, Inc. [AMD] Zeppelin USB 3.0 xHCI Compliant Host Controller
0f:00.0 Non-Essential Instrumentation [1300]: Advanced Micro Devices, Inc. [AMD] Zeppelin/Renoir PCIe Dummy Function
0f:00.2 SATA controller: Advanced Micro Devices, Inc. [AMD] FCH SATA Controller [AHCI mode] (rev 51)
0f:00.3 Audio device: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) HD Audio Controller
40:00.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Root Complex
40:00.2 IOMMU: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) I/O Memory Management Unit
40:01.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-1fh) PCIe Dummy Host Bridge
40:01.1 PCI bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) PCIe GPP Bridge
40:02.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-1fh) PCIe Dummy Host Bridge
40:03.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-1fh) PCIe Dummy Host Bridge
40:03.1 PCI bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) PCIe GPP Bridge
40:04.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-1fh) PCIe Dummy Host Bridge
40:07.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-1fh) PCIe Dummy Host Bridge
40:07.1 PCI bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Internal PCIe GPP Bridge 0 to Bus B
40:08.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-1fh) PCIe Dummy Host Bridge
40:08.1 PCI bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Internal PCIe GPP Bridge 0 to Bus B
41:00.0 Non-Volatile memory controller: SK hynix PC401 NVMe Solid State Drive 256GB
42:00.0 PCI bridge: Advanced Micro Devices, Inc. [AMD/ATI] Device 14a0 (rev 01)
43:00.0 PCI bridge: Advanced Micro Devices, Inc. [AMD/ATI] Device 14a1
44:00.0 Display controller: Advanced Micro Devices, Inc. [AMD/ATI] Vega 20 [Radeon Pro VII/Radeon Instinct MI50 32GB] (rev 01)
45:00.0 Non-Essential Instrumentation [1300]: Advanced Micro Devices, Inc. [AMD] Zeppelin/Raven/Raven2 PCIe Dummy Function
45:00.2 Encryption controller: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-0fh) Platform Security Processor (PSP) 3.0 Device
45:00.3 USB controller: Advanced Micro Devices, Inc. [AMD] Zeppelin USB 3.0 xHCI Compliant Host Controller
46:00.0 Non-Essential Instrumentation [1300]: Advanced Micro Devices, Inc. [AMD] Zeppelin/Renoir PCIe Dummy Function
46:00.2 SATA controller: Advanced Micro Devices, Inc. [AMD] FCH SATA Controller [AHCI mode] (rev 51)

Truenas fangtooth 25.04.2.4 up to date
running application : jellyfin (no GPU passtrough) // Nextcloud (no GPU passtrough) // ollama // onlyoffice-document-server // open-webui (no GPU passtrough)

StevenMcE · September 20, 2025, 2:08am

Tbh check if Debian 6.12 has drivers by default for it or if you have to install them. If you need to add them likely will have to use install-dev-tools to write to /usr

janot928 · September 20, 2025, 11:15am

I tried to use the install-dev tools, and it gave me this message :

> root@truenas[~]# install-dev-tools
> + FORCE_ARG=
> + [[ '' == \-\-\f\o\r\c\e ]]
> + [[ ! -S /var/run/middleware/middlewared.sock ]]
> + PACKAGES=(make open-iscsi python3-cryptography python3-pip python3-pyfakefs python3-pyotp python3-pytest python3-pytest-asyncio python3-pytest-dependency python3-pytest-rerunfailures python3-pytest-timeout snmp sshpass zstd)
> + PIP_PACKAGES=()
> + '[' -f /usr/local/libexec/disable-rootfs-protection ']'
> + /usr/local/libexec/disable-rootfs-protection
> /usr is currently provided by a readonly systemd system extension. This may occur if nvidia module support is enabled. System extensions must be disabled prior to disabling rootfs protection.

Could it be that the previous Nvidia GPU driver causes some interference with the amd drivers ? How do i remove them ? how can I install the amd driver when i cannot use dpkg ?

SmallBarky · September 20, 2025, 8:11pm

From Documents

Users with NVIDIA GPU drivers installed cannot enable developer mode while the NVIDIA kernel module is mounted. Running install-dev-tools in this state results in the following error:

/usr is currently provided by a readonly systemd system extension.
This may occur if nvidia module support is enabled. System extensions
must be disabled prior to disabling rootfs protection.

This happens because the NVIDIA drivers are overlaid onto /usr via systemd-sysext, making it read-only by design. To resolve the issue, unmerge systemd-sysext, run install-dev-tools, then merge system extensions again.

NexGen3D · November 3, 2025, 6:41am

How did you go with the MI50 card in TrueNAS, did it end up working okay using ROCM? Looking to pick a couple of these up as they are dirt cheap at the moment, would be interesting to see how you went.

janot928 · November 3, 2025, 5:48pm

Hi,

@NexGen3D In the end I made a dedicated machine for LLM. As far as I can tell, you cannot install ROCM drivers on truenas scale.

NexGen3D · November 4, 2025, 12:48am

I’m starting to think its a far better idea to move TrueNAS off bare metal and run it on Proxmox, I used to do this as in the past.

I’m only looking at an MI50 as I have been playing sround with LocalAI on TrueNAS, but if we can’t install ROCM or its not included by default, then its pretty useless, nVidia is not really supported but I thought AMD would be, I have a spare RX 6700 sitting here, I might throw that in and see what does and doesnt work brfore I purchase anything,

Was looking at a 32GB MI50 for LocalAI and use my existing Arc GPU for other tasks, I’m finding ROCM issues with just about every thing I try, not looking good to be honest.

Did you test the card with Vulkan at all?

Edit: I’ve tested Vulcan under 25.10 with a RX 6700 (RDNA 2) for LLM acceleration and its fully working, but ROCM is not, I’ve also found that there is no monitoring software on the system for AMD, like amdgpu_top, but intel_gpu_top is there, would be nice to see something either in the GUI or in the shell to properly monitor GPU’s.

So I’m on the fence with AMD GPU’s in TrueNAS, I’ll need to test a newer one to see how well supported that is, but I think, the future might be with a B50 or B60 workstation cards from Intel.

NexGen3D · November 18, 2025, 6:39am

If anyone else is interested, you don’t need to install any drivers for ROCM in TrueNAS to run Ollama, well not for RDNA 2 or 4, currently running a single RX 9060 XT 16GB and ROCM is fully functional, can’t speak for older cards like the Mi50 or Radeon VII class (Vega) but newer cards are fully working, you just need to select the ROCM version of Ollama when spinning up the app.