GPU CUDA problems: CUDA_ERROR_UNKNOWN

Hi I use a GeForce GTX 980 with Ubuntu 15.04, but I have trouble using the GPU from within Matlab. When i run gpuDevice I get the following error " >> g=gpuDevice Error using gpuDevice (line 26) An unexpected error occurred during CUDA execution. The CUDA error was: CUDA_ERROR_UNKNOWN "

Can anyone help me figuring out what the problem is?

[Joss Knight]

What version of MATLAB are you using?

Answer 1

[Ashish Gudla]

It is possible that the CUDA driver might be causing the error message. Do you have the latest drivers installed? If not, install the latest drivers and verify if you still see the error message.

You can get the drivers from:

http://www.nvidia.com/Download/index.aspx

[hongshaorou]

thanks,I am trying to solve.

Answer 2

[Wajahat]

Try starting matlab with sudo:

>> sudo /usr/local/MATLAB/R2014a/bin/matlab -desktop

Answer 3

[Douglas]

It seems to be a permission error. I've started matlab as root and, in my case, the problem was solved.

To run as normal user, see this:

https://bugs.launchpad.net/ubuntu/+source/nvidia-graphics-drivers-331/+bug/1361207

You have to install nvidia-modprobe package.

Answer 4

[Yue Bai]

I met the same problem, solved with restart computer

Answer 5

[SATYAM KUMAR CHOURASIYA 16MCA1038]

 Error using parallel.gpu.CUDADevice.hBuild
    An unexpected error occurred trying to retrieve CUDA device properties. The CUDA error was:
    CUDA_ERROR_UNKNOWN

my gpu device having computing capacity of 2.1 ,

can u elaborate this is the reason of this issue or something else?

Answer 6

[Yuya Kodama]

In my environment, everything from R2020b to R2016a had the same error with no gpu available.

The GPU is Tesla K80 * 2 and the OS is Ubuntu 18,04.

When I checked, nvidia-smi was OK, but even python tensorflow could not call the GPU.

The reason was that my predecessor had left behind a blacklist.conf configuration file with a fixed driver version.

So, all drivers could not be loaded by modprove.

Therefore, I think that the dependencies were broken with the version upgrade.

Answer 7

[bai yang]

I solved this problem with run as administrator. You can just have a try.