Here is the diagnostic log and also the full error message.

I couldn't find the forum post anymore, but remember I had the same issue with prior updates and was told that this is a structrual issue / bug.

--------------Can't download Nvidia Driver Package v580.105.08-----------------
Removing package: nvidia-driver-2025.03.25
plugin: run failed: '/bin/bash' returned 1
Executing hook script: CA_postHook
Clearing Community Applications plugin cache
Executing hook script: post_plugin_checks
Plugin installed

tower-diagnostics-20251121-2056.zip

40 minutes ago, steve1977 said:

Here is the diagnostic log and also the full error message.

Are you sure that nothong on your network is blocking access to the GitHub API, pleas check especially if you have Unifi network gear on your network or any kond of AdBlocking.

42 minutes ago, steve1977 said:

I couldn't find the forum post anymore, but remember I had the same issue with prior updates and was told that this is a structrual issue / bug.

That's probably old information but even if you have this kind of issue please report that since such an issue needs to be fixed.

However there is no such issue to my knowledge because other people would have reported that here too.

It somehow started working again. Maybe indeed a network issue.

I remember someone mentioned this being a known issue, but couldn't find where. Maybe discord. Great this should no longer be the case. Will report back if/once the issue comes back.

Hi, I’m running into a weird and very inconsistent issue with my RTX 5070 FE on Unraid.

On fresh boot, GPU Statistics can see the GPU normally, but as soon as anything tries to use the GPU (starting a Docker container, doing a CUDA task, or sometimes even when doing nothing), it quickly throws: Vendor command returned unparseable data.

Sometimes the GPU works for a while and I can do transcoding or ML tasks, but it’s extremely unstable. Restarting or stopping GPU-dependent containers may temporarily fix it.

I’m on Unraid 7.2.0 (same issue on 7.1.4), using the NVIDIA open-kernel driver (580.95.05).

I also tried switching to the proprietary driver (580.105.08), but then the GPU is not detected at all — the plugin shows “No GPU found”, and nvidia-smi says “No devices were found”.

HDMI/DP video output works normally. Unraid “System Devices” correctly lists the card as GeForce RTX 5070.

The GPU is an RTX 5070 Founders Edition 12GB. i5-11600KF with ASUS TUF B560M (PCIe 4.0). No iGPU on the CPU. PSU is Seasonic Focus ATX3 GX-850.

---

To rule out hardware issues, I tested on Windows Server 2025 with the latest Game Ready 581.80 driver, running OCCT VRAM and 3D Adaptive tests for one hour each.

Both tests were completely stable — temps and clocks rose to normal levels and stayed flat with no drops, errors, or instability.

This makes me think the hardware itself is probably fine.

---

Here are examples of the repeating Unraid syslog errors when “Vendor command returned unparseable data.” appears (open-kernel driver (580.95.05); only pasting a few; the logs spam continuously):

A:

Nov 26 19:02:16 Jonathan-Server kernel: NVRM: nvCheckOkFailedNoLog: Check failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from memdescAllocInternal(pMemDesc) @ memdesc.c:1359
Nov 26 19:02:16 Jonathan-Server kernel: NVRM: nvAssertOkFailedNoLog: Assertion failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from nvStatus @ message_queue_cpu.c:242
Nov 26 19:02:16 Jonathan-Server kernel: NVRM: _kgspInitRpcInfrastructure: GspMsgQueueInit failed
Nov 26 19:02:16 Jonathan-Server kernel: NVRM: kgspConstructEngine_IMPL: init RPC infrastructure failed
Nov 26 19:02:16 Jonathan-Server kernel: NVRM: osInitNvMapping: *** Cannot attach gpu
Nov 26 19:02:16 Jonathan-Server kernel: NVRM: RmInitAdapter: osInitNvMapping failed, bailing out of RmInitAdapter
Nov 26 19:02:16 Jonathan-Server kernel: NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x22:0x51:763)
Nov 26 19:02:16 Jonathan-Server kernel: NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0

B:

Nov 17 02:36:53 Jonathan-Server kernel: NVRM: nvCheckOkFailedNoLog: Check failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from _memdescAllocInternal(pMemDesc) @ mem_desc.c:1353
Nov 17 02:36:53 Jonathan-Server kernel: NVRM: nvAssertOkFailedNoLog: Assertion failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from status @ kernel_gsp.c:4615
Nov 17 02:36:53 Jonathan-Server kernel: NVRM: nvCheckOkFailedNoLog: Check failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from kgspCreateRadix3(pGpu, pKernelGsp, &pKernelGsp->pGspUCodeRadix3Descriptor, NULL, pGspFw->pImageData, pGspFw->imageSize) @ kernel_gsp.c:4502
Nov 17 02:36:53 Jonathan-Server kernel: NVRM: kgspInitRm_IMPL: Error preparing GSP-RM image
Nov 17 02:36:53 Jonathan-Server kernel: NVRM: RmInitAdapter: Cannot initialize GSP firmware RM
Nov 17 02:36:53 Jonathan-Server kernel: NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x62:0x51:1941)
Nov 17 02:36:53 Jonathan-Server kernel: NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0

Here is the syslog when using the proprietary driver (580.105.08) — the Blackwell 5070 needs open kernel modules:

Nov 27 10:38:24 Jonathan-Server kernel: NVRM: The NVIDIA GPU 0000:01:00.0 (PCI ID: 10de:2f04)
Nov 27 10:38:24 Jonathan-Server kernel: NVRM: installed in this system requires use of the NVIDIA open kernel modules.
Nov 27 10:38:24 Jonathan-Server kernel: NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x22:0x56:884)
Nov 27 10:38:24 Jonathan-Server kernel: NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0

---

Given the GPU has 12GB VRAM and nothing is even using it at the time, the “Out of memory [NV_ERR_NO_MEMORY]” part seems unlikely to be real VRAM exhaustion.

Is this likely a problem with the Linux open-kernel driver for the RTX 50-series? Or something with Unraid’s kernel?

Or could this still be some rare hardware or compatibility issue?

Appreciate any insights — thanks!

55 minutes ago, JonathanChiu said:

Given the GPU has 12GB VRAM and nothing is even using it at the time, the “Out of memory [NV_ERR_NO_MEMORY]” part seems unlikely to be real VRAM exhaustion.

This has nothing to do with your VRAM, this message usually means your System reserved memory which is a part of your RAM.

However can you try to add nvidia-persistenced to your go file and reboot please?

This is a common issue with Nvidia 50 series cards and you can find also much posts on the Nvidia Developer Forums.

Most of the times it seems like a hardware incompatibility issue or boils down to a configuration issue.

Even I encountered this on my Linux Desktop PC running Debian with a 30 series card and the Open Source Driver, the solution was to apply the default settings in my BIOS, I assume that it had something to do with ACPI on my machine.

However I would recommend that you try:

1. Updating the BIOS

2. Try it with the default BIOS settings

3. Lower the PCIe Gen to something like Gen3 or even Gen4

4. Make sure to enable Above 4G Decoding and Resizable BAR Support in your BIOS

1 hour ago, JonathanChiu said:

Is this likely a problem with the Linux open-kernel driver for the RTX 50-series? Or something with Unraid’s kernel?

More driver related than Kernel related but I'm not too sure, most of the times it's a hardware incompatibility issue, or a misconfiguration.

The message that you see is that the card can't be reset properly and therefore initialization fails.

In the future please always add your Diagnostics since I don't even know which hardware do you use (CPU, Motherboard,...) since most users who reported such issues are on AMD Systems.

15 hours ago, ich777 said:

This has nothing to do with your VRAM, this message usually means your System reserved memory which is a part of your RAM.

However can you try to add nvidia-persistenced to your go file and reboot please?

This is a common issue with Nvidia 50 series cards and you can find also much posts on the Nvidia Developer Forums.

Most of the times it seems like a hardware incompatibility issue or boils down to a configuration issue.

Even I encountered this on my Linux Desktop PC running Debian with a 30 series card and the Open Source Driver, the solution was to apply the default settings in my BIOS, I assume that it had something to do with ACPI on my machine.

However I would recommend that you try:

1. Updating the BIOS

2. Try it with the default BIOS settings

3. Lower the PCIe Gen to something like Gen3 or even Gen4

4. Make sure to enable Above 4G Decoding and Resizable BAR Support in your BIOS

More driver related than Kernel related but I'm not too sure, most of the times it's a hardware incompatibility issue, or a misconfiguration.

The message that you see is that the card can't be reset properly and therefore initialization fails.

In the future please always add your Diagnostics since I don't even know which hardware do you use (CPU, Motherboard,...) since most users who reported such issues are on AMD Systems.

I just tested the RTX 5070 on Ubuntu 24.04 with open-kernel driver 580.95.05, and everything works fine. I have done everything you recommended, but the problem still persists on Unraid. The following is the diagnostic file.

jonathan-server-diagnostics-20251128-1353.zip

41 minutes ago, JonathanChiu said:

I just tested the RTX 5070 on Ubuntu 24.04 with open-kernel driver 580.95.05, and everything works fine. I have done everything you recommended, but the problem still persists on Unraid. The following is the diagnostic file.

Sorry but I don't support pirated versions from Unraid:

/boot/config/go:

export UNRAID_GUID=0951-1666-DBDE-E74128760E31

export UNRAID_NAME=Jonathan-Server

export UNRAID_DATE=1654646400

export UNRAID_VERSION=Pro

/lib64/ld-linux-x86-64.so.2 /boot/config/unraider

@SpencerJ

Thank you for your help, I’m so sorry. If it’s a problem with your NVIDIA driver for Unraid, please fix it. Thanks!

1 minute ago, JonathanChiu said:

If it’s a problem with your NVIDIA driver for Unraid, please fix it. Thanks!

I would rather recommend that you buy a legit version from Unraid. Thanks!

13 hours ago, JonathanChiu said:

Thank you for your help, I’m so sorry. If it’s a problem with your NVIDIA driver for Unraid, please fix it. Thanks!

Buy a real license please.

Hi

Is there anyway I can install an older driver 474.82 for my old GT730; rather than the latest?

My use case : I don't really need the latest drivers for modern GPUs (I don't have the room or PCIe for them), and the GT730 fits into a PCI-E x4 slot.

I've tried pretty much everything to get a VM to work with it so suspect its the new driver causing QEMU failing errors.

Thanks

Chris

49 minutes ago, SmokeyColes said:

Is there anyway I can install an older driver 474.82 for my old GT730; rather than the latest?

I assume you are on 7.2.2, if yes, the answer is no because the legacy driver doesn't compile against newer Kernels and is basically deprecated.

51 minutes ago, SmokeyColes said:

I've tried pretty much everything to get a VM to work with it so suspect its the new driver causing QEMU failing errors.

You don't need the Nvidia driver if you want to use it in a VM, please read the first post again.

It is not recommended to install the driver at all when you want to use the card in a VM.

@ich777 apologies

So I'm pretty sure I messed up.

I have been running a p2000 for a while with no issues. I got a 5060ti the other day and was getting this issue. NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver.

I got the v580.95.05 open source driver but was still getting the issue. I restarted a few times, change pcie to gen 4, updated bios, uninstalled and reinstalled the plugin, and updated unraid from 7.0 to 7.2.2.

The error then changed to something like No devices were found. I tried restarting, then I tried running "nvidia-smi -pm 1" from the xploit61 post on November 2nd. I can't remember if I ran this one as well or not "nvidia-persistenced" from ich777 post on November 2nd. However, now when I try to reboot it goes to the boot gui and then the display is gets no signal and unraid does not launch.

I am able to get unraid to launch with my p2000 or with my p2000 and 5060ti. when I have both in they now show up for whatever reason. but I can't get it to boot with just the 5060ti.

attached is a screen shot of the plugin page and 'nvidia-smi' output. I am unsure how to pull server diagnostics without ip and info at this moment.

thanks for any and all help in advance!

6 hours ago, killercola said:

So I'm pretty sure I messed up.

Do you have Diagnostics?

6 hours ago, killercola said:

then I tried running "nvidia-smi -pm 1" from the xploit61 post on November 2nd

Please don't, as mentioned a few posts or even one post below this is not the preferred method, simply use:
nvidia-persistenced

6 hours ago, killercola said:

attached is a screen shot of the plugin page and 'nvidia-smi' output. I am unsure how to pull server diagnostics without ip and info at this moment.

What do you mean by that? How did you get that screenshot then?

You can always create Diagnostics when opening a Terminal or even locally and type in: diagnostics

This will then be saved to the logs directory on your USB Boot device.

diagnostics attached.

Edited December 6, 2025Dec 6 by killercola

23 hours ago, killercola said:

However, now when I try to reboot it goes to the boot gui and then the display is gets no signal and unraid does not launch.

Are you sure that it doesn't boot or do you see just a blinking cursor in the top left corner?

23 hours ago, killercola said:

I am able to get unraid to launch with my p2000 or with my p2000 and 5060ti. when I have both in they now show up for whatever reason. but I can't get it to boot with just the 5060ti.

Just to be clear, the OpenSource driver only supports GPUs RTX20 and up (including the GTX16 series) and the other drivers only supports up to RTX40 series.

So to speak you can't utilize both cards in your system, it's either you P2000 or your RTX5060ti.

I'm not sure what's going on with the boot issue since this seems very strange to me and usually has nothing to do with the Nvidia Driver.

From what I see from you Diagnostics, the RTX5060ti is properly recognized and should work, I see no other obvious issue.

If the card is dropping from the Bus then you can try to put nvidia-persistenced in your go file. Please also check that your Power Supply is up to the task since they degrade over time.

it goes to the boot gui and then the display says no signal and turns off. in about 5 min it will try again.

when using unraid 7.0 it would not detect the gpu driver. I'm going to try rolling the os back to 7.0 to see if that might fix things.

my intent is to only use the 5060 but until I can solve the boot issue I will have to keep the p2000

thanks for helping with what you can!

Edited December 3, 2025Dec 3 by killercola

ok looks like the boot issue was related to 7.2.2

but now I can't get it to detect the gpu.

uploading the new images and diagnostics. is it possible I just have a defective gpu?

Edited December 6, 2025Dec 6 by killercola

21 hours ago, killercola said:

but now I can't get it to detect the gpu.

Are you sure that Above 4G Decoding is enabled and Resizable BAR Support is enabled? You have a lot of these messages in your sysslog:

Dec 3 08:48:41 Tower kernel: NVRM: This PCI I/O region assigned to your NVIDIA device is invalid:

Dec 3 08:48:41 Tower kernel: NVRM: BAR1 is 0M @ 0x0 (PCI:0000:0c:00.0)

Dec 3 08:48:41 Tower kernel: NVRM: This PCI I/O region assigned to your NVIDIA device is invalid:

Dec 3 08:48:41 Tower kernel: NVRM: BAR2 is 0M @ 0x0 (PCI:0000:0c:00.0)

and as a result from this message:

Dec 3 08:50:36 Tower kernel: NVRM: kbusVerifyBar2_GB202: MMUTest BAR0 window offset 0x70e000 returned garbage 0x0

Dec 3 08:50:36 Tower kernel: NVRM: nvAssertOkFailedNoLog: Assertion failed: Generic memory error [NV_ERR_MEMORY_ERROR] (0x00000072) returned from kbusVerifyBar2_HAL(pGpu, pKernelBus, NULL, NULL, 0, 0) @ kern_bus_gm107.c:352

Dec 3 08:50:36 Tower kernel: NVRM: nvAssertOkFailedNoLog: Assertion failed: Generic memory error [NV_ERR_MEMORY_ERROR] (0x00000072) returned from kbusStateInitLockedKernel_HAL(pGpu, pKernelBus) @ kern_bus_gm107.c:457

Dec 3 08:50:36 Tower kernel: NVRM: RmInitNvDevice: *** Cannot initialize the device

Dec 3 08:50:36 Tower kernel: NVRM: RmInitAdapter: RmInitNvDevice failed, bailing out of RmInitAdapter

Dec 3 08:50:36 Tower kernel: NVRM: rmapiReportLeakedDevices: Device object leak: (0xc1e00003, 0xcaf00000). Please file a bug against RM-core.

Dec 3 08:50:36 Tower kernel: NVRM: nvAssertFailedNoLog: Assertion failed: 0 @ rmapi.c:935

Dec 3 08:50:36 Tower kernel: NVRM: rmapiReportLeakedDevices: Device object leak: (0xc1e00004, 0xcaf00000). Please file a bug against RM-core.

Dec 3 08:50:36 Tower kernel: NVRM: nvAssertFailedNoLog: Assertion failed: 0 @ rmapi.c:935

Dec 3 08:50:36 Tower kernel: NVRM: nvAssertFailedNoLog: Assertion failed: listCount(&pKernelBus->virtualBar2[gfid].usedMapList) == 0 @ kern_bus_vbar2.c:346

Dec 3 08:50:36 Tower kernel: NVOC: __nvoc_objDelete: Child class KernelVideoEngine not freed from parent class OBJGPU.NVRM: iovaspaceDestruct_IMPL: 4 left-over mappings in IOVAS 0xc00

Dec 3 08:50:36 Tower kernel: NVRM: GPU 0000:0c:00.0: RmInitAdapter failed! (0x24:0x72:1100)

Dec 3 08:50:36 Tower kernel: NVRM: GPU 0000:0c:00.0: rm_init_adapter failed, device minor number 0

Dec 3 08:50:36 Tower kernel: NVRM: _kgspBootGspRm: unexpected WPR2 already up, cannot proceed with booting GSP

Dec 3 08:50:36 Tower kernel: NVRM: _kgspBootGspRm: (the GPU is likely in a bad state and may need to be reset)

Dec 3 08:50:36 Tower kernel: NVRM: RmInitAdapter: Cannot initialize GSP firmware RM

Dec 3 08:50:36 Tower kernel: NVRM: GPU 0000:0c:00.0: RmInitAdapter failed! (0x62:0x40:1860)

Dec 3 08:50:36 Tower kernel: NVRM: GPU 0000:0c:00.0: rm_init_adapter failed, device minor number 0

Hello everyone,

I'm having an issue with the latest driver version:

I tried to revert back to the older version, which was running (580.105.08), and my TeslaP4 is back up and running.

Is there some problem with the latest driver or what?

2 hours ago, TDA said:

I tried to revert back to the older version, which was running (580.105.08), and my TeslaP4 is back up and running.

Thank you for this. I updated my server to 7.2.2 yesterday and jellyfin kept failing saying it couldn't see the card. I tried changing the driver to the latest, and the open source and neither worked for me. After changing to the 580.105.08 driver, my 1050ti started working again.

9 hours ago, ich777 said:

Are you sure that Above 4G Decoding is enabled and Resizable BAR Support is enabled? You have a lot of these messages in your sysslog:

THANK YOU. That was the issue....

9 hours ago, Taubin said:

I tried changing the driver to the latest, and the open source and neither worked for me. After changing to the 580.105.08 driver, my 1050ti started working again.

11 hours ago, TDA said:

Is there some problem with the latest driver or what?

First of all, does anyone of you two have Diagnostics from where the issue occurred?

Otherwise it is really hard to troubleshoot.

You are talking about Driver version 590.44.01 correct (please also do keep in mind that this is a beta driver - latest is not always grates :D )?

Issue here is more the hardware that you are using, this is the first driver which is dropping everything before the first RTX series (Turing), so to speak Pascal and everything before is not supported anymore, please see (under the section Supported Products) :

https://www.nvidia.com/en-us/drivers/details/258750/

You now have to make sure that you stay below driver version 590

Hello-

I am also getting the "NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running" error after upgrading my driver to 590.44.01 this morning. Any docker that uses the extra parameters "--runtime=nvidia" will no longer start. I've attached my diagnostics to hopefully troubleshoot the issue.

Thank you!

tower-diagnostics-20251205-0740.zip