The installation described here was done the 10/20/2016 on a freshen udpate system by using :
We add the nvidia key and repository from www.elrepo.org:
sudo rpm --import https://www.elrepo.org/RPM-GPG-KEY-elrepo.org sudo rpm -Uvh http://www.elrepo.org/elrepo-release-7.0-2.el7.elrepo.noarch.rpm
Then we install nvidia-detect
When we run nvidia-detect, we obtain as result kmod-nvidia. So we install it
The version of the installed driver is the 367.57-1. A reboot is needed to use the new driver.For each CUDA version there is a minimal version of Nvidia driver for which CUDA functionality works
CUDA version | Driver version | download link |
6.5 | >= 340.00 | cuda_6.5.14_linux_64.run |
7.0 | >= 346.00 | cuda_7.0.28_with_346.75_linux.run |
7.5 | >= 352.00 | cuda_7.5.18_linux.run |
8.0 | >= 361.00 | cuda_8.0.44_linux-run |
So we can download and install all these versions!
Before starting CUDA installation, we install the gcc, g++ and gfortran compiler
The current version is the 4.8.5-4.We also install some libraries to avoid some warning messages
Otherwise, on my computer, I have these warning messages during the installation process:
Due to an unsupported compiler (version 4.8.5), we use –override option for CUDA 6.5 installation.
chmod u+x cuda_*.run sudo ./cuda_6.5.14_linux_64.run --silent --verbose --toolkit --samples --samplespath=/usr/local/cuda-6.5/samples --override sudo ./cuda_7.0.28_with_346.75_linux.run --silent --verbose --toolkit --samples --samplespath=/usr/local/cuda-7.0/samples sudo ./cuda_7.5.18_linux.run --silent --verbose --toolkit --samples --samplespath=/usr/local/cuda-7.5/samples sudo ./cuda_8.0.44_linux-run --silent --verbose --toolkit --samples --samplespath=/usr/local/cuda-8.0/samples