I installed Cuda 12.0 on a fresh install of Ubuntu 22.04. Following instructions, including the verification suggested installation was successful. I cloned Cuda git repo and built it. Nvidia-smi reports correctly I have 3070 TI installed,
±----------------------------------------------------------------------------+
| NVIDIA-SMI 525.85.12 Driver Version: 525.85.12 CUDA Version: 12.0 |
|-------------------------------±---------------------±---------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 NVIDIA GeForce … On | 00000000:01:00.0 On | N/A |
| 0% 35C P8 11W / 290W | 778MiB / 8192MiB | 0% Default |
| | | N/A |
±------------------------------±---------------------±---------------------+
±----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| 0 N/A N/A 2259 G /usr/lib/xorg/Xorg 382MiB |
| 0 N/A N/A 2723 G …ome-remote-desktop-daemon 3MiB |
| 0 N/A N/A 2764 G /usr/bin/gnome-shell 73MiB |
| 0 N/A N/A 5113 G …050822919761424938,131072 88MiB |
| 0 N/A N/A 5357 G …5/usr/lib/firefox/firefox 60MiB |
| 0 N/A N/A 47735 G …b/thunderbird/thunderbird 166MiB |
±----------------------------------------------------------------------------+
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2023 NVIDIA Corporation
Built on Fri_Jan__6_16:45:21_PST_2023
Cuda compilation tools, release 12.0, V12.0.140
Build cuda_12.0.r12.0/compiler.32267302_0
Not all examples are build … ex. I cannot find deviceQuery so I used one from extras/demo_suite.
the outcome is
/usr/local/cuda-12.0/extras/demo_suite$ ./deviceQuery
./deviceQuery Starting…
CUDA Device Query (Runtime API) version (CUDART static linking)
cudaGetDeviceCount returned 100
→ no CUDA-capable device is detected
Result = FAIL
Based on the installation suggestions this can have something to do with permissions… after checking
crw-rw-rw- 1 root root 195, 0 Mar 2 13:08 /dev/nvidia0
crw-rw-rw- 1 root root 195, 255 Mar 2 13:08 /dev/nvidiactl
crw-rw-rw- 1 root root 195, 254 Mar 2 13:08 /dev/nvidia-modeset
crw-rw-rw- 1 root root 509, 0 Mar 2 13:08 /dev/nvidia-uvm
crw-rw-rw- 1 root root 509, 1 Mar 2 13:08 /dev/nvidia-uvm-tools
it seems they are correctly set to 0666 … as noted in the installation docs.
Can anybody suggest possible recovery path?