Where did CUDA get installed in my computer?

45
votes

I'm trying to install CUDA 7.5 in my ubuntu 14.04. I followed everything in this guide (installation through package): http://developer.download.nvidia.com/compute/cuda/7.5/Prod/docs/sidebar/CUDA_Installation_Guide_Linux.pdf until post build section. Mainly, by running the following commands: 

sudo dpkg -i cuda-repo-ubuntu1404-7-5-local_7.5-18_amd64.deb
sudo apt-get update
sudo apt-get install cuda

The thing is the folder /usr/local/cuda* does not exist after successful CUDA installation. Further trying to install cuda says that it is already the newest version.

sudo apt-get install cuda
Reading package lists... Done
Building dependency tree       
Reading state information... Done
cuda is already the newest version.

ls /usr/local/cuda*
ls: cannot access /usr/local/cuda*: No such file or directory

It also does not find nvcc.

nvcc
The program 'nvcc' is currently not installed. You can install it by typing:
sudo apt-get install nvidia-cuda-toolkit

sudo find /usr/ -name nvcc
<no output>

What is wrong?

4
cudainstallation
Try sudo find / -name nvcc. If that turns up nothing, then the simple fact is that CUDA did not get installed on your computer. In that case, I would suggest starting over with a clean OS load of Ubuntu 14.04 and repeat the install steps from the guide you linked. - Robert Crovella
dpkg -L cuda will show you where all the files installed by that package are located. This is much more a Ubuntu package management question than anything to do with CUDA, and I have voted to close it as such - talonmies
@RobertCrovella, A clean OS load of Ubuntu? are you kidding me? - krips89
@talonmies, I think it is something with the package of cuda. The latest one probably do not have cuda toolkit files in them. Trying to install through runfile as described in the answer solved the problem. - krips89

4 Answers

42
votes

Usually, it is /usr/local/cuda. If this is not the case, you can try to locate cuda. If you want to find directories only, run

locate cuda | grep /cuda$

or

find / -type d -name cuda 2>/dev/null

For me, it turned out to be in /opt/cuda-7.5

14
votes

I solved (ditched actually) the problem by using 'Runfile method' for installing. I could get the latest nvidia driver installed with the package method explained above, but the problem seemed to be the cuda toolkit.

Installing the driver through Runfile is pain. So, at the prompt I chose to install just 'cuda toolkit 7.5' and all the files got copied to /usr/local/cuda* directory properly.

9
votes

I tracked the CUDA installation folder to /usr/lib/nvidia-cuda-toolkit. How? I used locate nvcc.

I had installed NVIDIA driver using the Software and Updates --> Additional Drivers followed by CUDA Toolkit installation using sudo apt install nvidia-cuda-toolkit on Ubuntu 20.04. nvcc --version was working fine but when it came to verifying cuDNN installation (https://docs.nvidia.com/deeplearning/cudnn/install-guide/index.html#verify), it was looking for the usr/local/cuda folder and failed.

I, therefore, created a symlink usr/local/cuda to /usr/lib/nvidia-cuda-toolkit as so: ln -s /usr/lib/nvidia-cuda-toolkit/ /usr/local/cuda.

Also added /usr/lib/nvidia-cuda-toolkit/libdevice to $LD_LIBRARY_PATH and /usr/lib/nvidia-cuda-toolkit/bin to$PATH variables.

All is working fine now.

0
votes

I had the same issue when upgrading to cuda 8.0. I solved it by changing the nvidia driver back to X.Org and then reinstall it from software& updates. You might want to delete old cuda files as well. I was able to reinstall cuda correctly after this.