Cannot query NVIDIA drivers on Ubuntu

Hello,

After having installed cuDNN + CUDA, I can’t query NVIDIA drivers.
Here is what I see while executing nvidia-smi:
NVIDIA-SMI has failed because it couldn’t communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running.

Here is what I see while executing dpkg -l | grep nvidia:

ii  libnvidia-cfg1-418:amd64                           418.87.00-0ubuntu1                               amd64        NVIDIA binary OpenGL/GLX configuration library
ii  libnvidia-common-418                               418.87.00-0ubuntu1                               all          Shared files used by the NVIDIA libraries
rc  libnvidia-compute-390:amd64                        390.116-0ubuntu0.18.04.1                         amd64        NVIDIA libcompute package
rc  libnvidia-compute-390:i386                         390.116-0ubuntu0.18.04.1                         i386         NVIDIA libcompute package
ii  libnvidia-compute-418:amd64                        418.87.00-0ubuntu1                               amd64        NVIDIA libcompute package
ii  libnvidia-decode-418:amd64                         418.87.00-0ubuntu1                               amd64        NVIDIA Video Decoding runtime libraries
ii  libnvidia-encode-418:amd64                         418.87.00-0ubuntu1                               amd64        NVENC Video Encoding runtime library
ii  libnvidia-fbc1-418:amd64                           418.87.00-0ubuntu1                               amd64        NVIDIA OpenGL-based Framebuffer Capture runtime library
ii  libnvidia-gl-418:amd64                             418.87.00-0ubuntu1                               amd64        NVIDIA OpenGL/GLX/EGL/GLES GLVND libraries and Vulkan ICD
ii  libnvidia-ifr1-418:amd64                           418.87.00-0ubuntu1                               amd64        NVIDIA OpenGL-based Inband Frame Readback runtime library
rc  nvidia-compute-utils-390                           390.116-0ubuntu0.18.04.1                         amd64        NVIDIA compute utilities
ii  nvidia-compute-utils-418                           418.87.00-0ubuntu1                               amd64        NVIDIA compute utilities
rc  nvidia-dkms-390                                    390.116-0ubuntu0.18.04.1                         amd64        NVIDIA DKMS package
ii  nvidia-dkms-418                                    418.87.00-0ubuntu1                               amd64        NVIDIA DKMS package
ii  nvidia-driver-418                                  418.87.00-0ubuntu1                               amd64        NVIDIA driver metapackage
rc  nvidia-kernel-common-390                           390.116-0ubuntu0.18.04.1                         amd64        Shared files used with the kernel module
ii  nvidia-kernel-common-418                           418.87.00-0ubuntu1                               amd64        Shared files used with the kernel module
ii  nvidia-kernel-source-418                           418.87.00-0ubuntu1                               amd64        NVIDIA kernel source package
ii  nvidia-modprobe                                    418.87.00-0ubuntu1                               amd64        Load the NVIDIA kernel driver and create device files
ii  nvidia-prime                                       0.8.8.2                                          all          Tools to enable NVIDIA's Prime
ii  nvidia-settings                                    418.87.00-0ubuntu1                               amd64        Tool for configuring the NVIDIA graphics driver
ii  nvidia-utils-418                                   418.87.00-0ubuntu1                               amd64        NVIDIA driver support binaries
ii  xserver-xorg-video-nvidia-418                      418.87.00-0ubuntu1                               amd64        NVIDIA binary Xorg driver

Any suggestions ?
Thanks.

Hi,

It seems you might be using older NVIDIA driver version.
Please refer to below support matrix to check if CUDA + cuDNN version installed on your system has compatible NVIDIA driver:
https://docs.nvidia.com/deeplearning/sdk/cudnn-support-matrix/index.html#cudnn-cuda-hardware-versions

Thanks

I didn’t find my card in the matrix from the link you provided.
Here is my NVIDIA board on Ubuntu machine:
~$ lspci -v | grep 3D
01:00.0 3D controller: NVIDIA Corporation GP107M [GeForce GTX 1050 Mobile] (rev a1)

Hi,

Please refer below link for check cuda compute capability of your GPU:
https://developer.nvidia.com/cuda-gpus

Thanks

Hi,

I’ve found my GPU under “GeForce and TITAN Products”, but haven’t found it under “GeForce Notebook Products”.
Does is mean that my GPU isn’t compatible with CUDA ?
Moreover, I haven’t found it on driver download page https://www.nvidia.com/download/index.aspx?lang=en-us

Hi,

Can you try installing/updating using GUI mode mentioned in below link:
https://www.linuxbabe.com/ubuntu/install-nvidia-driver-ubuntu-18-04

Thanks

Hi,

I’ve tried the approach on your link.
There is some ambiguity.
Here is overview of my machine:

And here is Additional Drivers tab:


How can I understand this: Nvidia driver is selected, but isn’t taken into account ?

Hi,

If you are using CUDA 10.2, you need to install NVIDIA driver 440.
Use below link for support matrix:
https://docs.nvidia.com/deeplearning/sdk/cudnn-support-matrix/index.html#cudnn-cuda-hardware-versions

Thanks

Hi,

Here is driver 440 download page:
https://www.nvidia.com/Download/driverResults.aspx/152647/en-us

Doesn’t work for me for 2 reasons:

  • there is no version for Linux
  • I don't see my graphical card (GE Force GTX 1050 Mobile) in the list of supported cards

Can you try installing the driver using above approach?
If it’s still not supported, then based on the support matrix you have to use CUDA 10.1.243.

Thanks

Hi,

As you can see on the screenshot from one of my previous messages there are 6 different NVIDIA drivers that are apparently available on my machine (390, 410, 415, 418, 435, 440).
Actually 418 is selected and it isn’t recognized … as we can see on “About” screenshot
Should I select 440 ?

Thanks.

Yes!
When I selected 440 it works.
Thanks.
What does the list of other drivers (i.e. 390 … 435) mean ?
Are they all installed on my machine.
If yes, should I remove them ?

As per the above output it seems that 418 driver is already installed on your system. You can either try v440 driver version to use latest CUDA version or install CUDA 10.1.243

Thanks

As per the above output it seems that 418 driver is already installed on your system. You can either try v440 driver version to use latest CUDA version or install CUDA 10.1.243

Yes, 418 and it seem 390 es well. Should I desinstall them ?
If yes, how ?