RTX 5060 Ti not working on Ubuntu 24.04 – All drivers (560/575/580) fail with “Failed to allocate NvkmsKapiDevice”

My new ZOTAC RTX 5060 Ti (Device ID 20d4) works perfectly in Windows, but on Ubuntu 24.04.3 LTS it never initializes with any NVIDIA driver.

System:

  • MSI PRO B760M-A WiFi

  • Intel 12th Gen CPU

  • Secure Boot: Disabled

  • BIOS Mode: UEFI

  • PEG/PCI_E1 Gen Mode: Gen4

  • Resizable BAR: Enabled

  • Display connected via DP to GPU

Ubuntu kernels tested:

  • 6.8.0-88-generic

  • 6.11 / 6.14 OEM kernels

Drivers tested (all fail):

  • 560.xx → repo install fails or GPU not detected

  • 575.xx → installs but GPU not detected

  • 580.119.02 (.run from NVIDIA) → modules load but GPU does not initialize

Consistent error across all drivers:


nvidia-drm: Failed to allocate NvkmsKapiDevice (GPU ID 0x0000:01:00)
nvidia-smi: No devices were found

Other observations:

  • lspci correctly shows the GPU and NVIDIA audio device

  • Nouveau is blacklisted

  • All NVIDIA kernel modules load (nvidia, nvidia-drm, nvidia-modeset, nvidia-uvm)

  • DKMS fails on OEM kernels, works on 6.8 but GPU still offline

  • BIOS set to PEG only, iGPU disabled / auto tested

Questions:

  1. Is RTX 5060 Ti (Device ID 20d4) fully supported by driver 560/575/580?

  2. Is kernel 6.8 supported for this GPU?

  3. Is this a known issue (KMS allocation failure) for RTX 50-series on Linux?

  4. Is a VBIOS update or upcoming driver required?

I can provide full logs if needed (dmesg, installer log, debug dumps).

Thank you.

Welcome @parul238 to the NVIDIA developer forums.

Please check »»»»»»»»»» If you have a problem, PLEASE read this first «««««««««« and provide the details mentioned there.

Thanks!

Also have this issue:
Motherboard: MSI MAG Z890 TOMAHAWK WIFI
CPU: Intel Core Ultra 7 270K Plus
GPU: MSI GeForce RTX 5060 Ti 16G GAMING OC
OS: Ubuntu 26.04 (Wayland)

Confirming that the system works perfectly on kernel 7.0.0-31 using the 595.84 driver line. However, updating to any newer kernel (tested on 7.0.0-38 and 7.0.0-42) causes an immediate boot crash with unreadable horizontal lines across the display.

Checked the historical journalctl logs of the broken boot sessions, and they confirm the identical error:
[drm:nv_drm_dev_load [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000200] Failed to allocate NvKmsKapiDevice

Compiling the newer 595.99.02 driver specifically into the 7.0.0-42 kernel does not resolve the issue; the memory allocation error persists. Pinning the system to the .31 kernel is currently the only workaround to keep a functioning display.

Please share nvidia-bug-report to check this issue further.

I removed some of the attachments as I received a bounce reply too many attachments.

(attachments)

debug_42.txt (1.78 KB)
nvidia-bug-report.log_2_31.gz (520 KB)
nvidia-bug-report.log_2_42_extradata.gz (287 KB)

Update:

AI and I finally arrived at a working system this morning.

It came down to getting the kernel and driver in sync.

Still not sure how they got out of sync but things are working now.
cire@cire-MS-7E32:~$ uname -a
Linux cire-MS-7E32 7.0.0-43-generic #43-Ubuntu SMP PREEMPT_DYNAMIC Fri Oct 23 13:25:25 UTC 2026 x86_64 GNU/Linux
cire@cire-MS-7E32:~$ cat /etc/os-release
PRETTY_NAME=“Ubuntu 26.04.1 LTS”
NAME=“Ubuntu”
VERSION_ID=“26.04”
VERSION=“26.04.1 LTS (Resolute Raccoon)”
VERSION_CODENAME=resolute
ID=ubuntu
ID_LIKE=debian
HOME_URL=“https://www.ubuntu.com/”
SUPPORT_URL=“https://help.ubuntu.com/”
BUG_REPORT_URL=“https://bugs.launchpad.net/ubuntu/”
PRIVACY_POLICY_URL=“https://www.ubuntu.com/legal/terms-and-policies/privacy-policy”
UBUNTU_CODENAME=resolute
LOGO=ubuntu-logo
cire@cire-MS-7E32:~$ nvidia-smi
Tue Nov 10 14:42:49 2026