The CUDA development environment relies on tight integration with the host development environment, including the host compiler and C runtime libraries, and is therefore only supported on distribution versions that have been qualified for this CUDA Toolkit release.

* Only Ubuntu 12.04 on x86 supports armhf cross development.

This document is intended for readers familiar with the Linux environment and the compilation of C programs from the command line. You do not need previous experience with CUDA or experience with parallel computation. Note: This guide covers installation only on systems with X Windows installed.

To verify that your GPU is CUDA-capable, go to your distribution's equivalent of System Properties, or, from the command line, enter:

lspci | grep -i nvidia

If you do not see any settings, update the PCI hardware database that Linux maintains by entering update-pciids (generally found in /sbin) at the command line and rerun the previous lspci command.

If your graphics card is from NVIDIA and it is listed in http://developer.nvidia.com/cuda-gpus, your GPU is CUDA-capable.

The Release Notes for the CUDA Toolkit also contain a list of supported products.

The CUDA Development Tools are only supported on some specific distributions of Linux. These are listed in the CUDA Toolkit release notes.

To determine which distribution and release number you're running, type the following at the command line:

uname -m && cat /etc/*release

You should see output similar to the following, modified for your particular system:

i386 Red Hat Enterprise Linux WS release 4 (Nahant Update 6)

The i386 line indicates you are running on a 32-bit system. On 64-bit systems running in 64-bit mode, this line will generally read: x86_64. The second line gives the version number of the operating system.

The gcc compiler and toolchain generally are installed as part of the Linux installation, and in most cases the version of gcc installed with a supported version of Linux will work correctly.

To verify the version of gcc installed on your system, type the following on the command line:

gcc --version

If an error message displays, you need to install the development tools from your Linux distribution or obtain a version of gcc and its accompanying toolchain from the Web.

For ARMv7 cross development, a suitable cross compiler is required. For example, performing the following on Ubuntu 12.04:

sudo apt-get install g++-4.6-arm-linux-gnueabihf

will install the gcc 4.6 cross compiler on your system which will be used by nvcc. Please refer to th NVCC manual on how to use nvcc to cross compile to the ARMv7 architecture

The CUDA Toolkit can be installed using either of two different installation mechanisms: distribution-specific packages, or a distribution-independent package. The distribution-independent package has the advantage of working across a wider set of Linux distributions, but does not update the distribution's native package management system. The distribution-specific packages interface with the distribution's native package management system. It is recommended to use the distribution-specific packages, where possible.

The NVIDIA CUDA Toolkit is available at http://developer.nvidia.com/cuda-downloads.

Choose the platform you are using and download the NVIDIA CUDA Toolkit

The CUDA Toolkit contains the CUDA driver and tools needed to create, build and run a CUDA application as well as libraries, header files, CUDA samples source code, and other resources.

$ md5sum <file>

• The package manager installations (RPM/DEB packages) and the stand-alone installer installations (.run file) of the NVIDIA driver are incompatible. Before using the distribution-specific packages, uninstall the NVIDIA driver with the following command:

  sudo /usr/bin/nvidia-uninstall

• For x86 to ARMv7 cross development, the host system must be an Ubuntu 12.04 system.

1. Perform the pre-installation actions.
2. Satisfy DKMS dependency

   The NVIDIA driver RPM packages depend on other external packages, such as DKMS and libvdpau. Those packages are only available on third-party repositories, such as EPEL. Any such third-party repositories must be added to the package manager repository database before installing the NVIDIA driver RPM packages, or missing dependencies will prevent the installation from proceeding.

3. Address custom xorg.conf, if applicable

   The driver relies on an automatically generated xorg.conf file at /etc/X11/xorg.conf. If a custom-built xorg.conf file is present, this functionality will be disabled and the driver may not work. You can try removing the existing xorg.conf file, or adding the contents of /etc/X11/xorg.conf.d/00-nvidia.conf to the xorg.conf file. The xorg.conf file will most likely need manual tweaking for systems with a non-trivial GPU configuration.

4. Install repository meta-data

   $ sudo rpm --install cuda-repo-<distro>-<version>.<architecture>.rpm

5. Clean Yum repository cache

   $ sudo yum clean expire-cache

6. Install CUDA

   $ sudo yum install cuda

   If the i686 libvdpau package dependency fails to install, try using the following steps to fix the issue:

   $ yumdownloader libvdpau.i686
   $ sudo rpm -U --oldpackage libvdpau*.rpm

7. Add libcuda.so symbolic link, if necessary

   The libcuda.so library is installed in the /usr/lib{,64}/nvidia directory. For pre-existing projects which use libcuda.so, it may be useful to add a symbolic link from libcuda.so in the /usr/lib{,64} directory.

8. Perform the post-installation actions.

1. Perform the pre-installation actions.
2. Address custom xorg.conf, if applicable

   The driver relies on an automatically generated xorg.conf file at /etc/X11/xorg.conf. If a custom-built xorg.conf file is present, this functionality will be disabled and the driver may not work. You can try removing the existing xorg.conf file, or adding the contents of /etc/X11/xorg.conf.d/00-nvidia.conf to the xorg.conf file. The xorg.conf file will most likely need manual tweaking for systems with a non-trivial GPU configuration.

3. Install repository meta-data

   $ sudo rpm --install cuda-repo-<distro>-<version>.<architecture>.rpm

4. Clean Yum repository cache

   $ sudo yum clean expire-cache

5. Install CUDA

   $ sudo yum install cuda

   If the i686 libvdpau package dependency fails to install, try using the following steps to fix the issue:

   $ yumdownloader libvdpau.i686
   $ sudo rpm -U --oldpackage libvdpau*.rpm

6. Add libcuda.so symbolic link, if necessary

   The libcuda.so library is installed in the /usr/lib{,64}/nvidia directory. For pre-existing projects which use libcuda.so, it may be useful to add a symbolic link from libcuda.so in the /usr/lib{,64} directory.

7. Perform the post-installation actions.

1. Perform the pre-installation actions.
2. Install repository meta-data

   $ sudo rpm --install cuda-repo-<distro>-<version>.<architecture>.rpm

3. Refresh Zypper repository cache

   $ sudo zypper refresh

4. Install CUDA

   $ sudo zypper install cuda

   The driver is provided in multiple packages, nvidia-gfxG03-kmp-desktop, nvidia-gfxG03-kmp-default, nvidia-gfxG03-kmp-trace, and their Unified Memory variants. When installing cuda, the correct driver packages should also be specified. Without doing this, zypper will select packages that may not work on the system. Run the following to detect the flavor of kernel and install cuda with the appropriate driver packages:

   $ uname -r
     3.4.6-2.10-<flavor>
   $ sudo zypper install cuda nvidia-gfxG03-kmp-<flavor> \
                         nvidia-uvm-gfxG03-kmp-<flavor>

5. Add the user to the video group

   $ sudo usermod -a -G video <username>

6. Install CUDA Samples GL dependencies

   The CUDA Samples package on SLES does not include dependencies on GL and X11 libraries as these are provided in the SLES SDK. These packages must be installed separately, depending on which samples you want to use.

7. Perform the post-installation actions.

1. Perform the pre-installation actions.
2. Install repository meta-data

   $ sudo rpm --install cuda-repo-<distro>-<version>.<architecture>.rpm

3. Refresh Zypper repository cache

   $ sudo zypper refresh

4. Install CUDA

   $ sudo zypper install cuda

   The driver is provided in multiple packages, nvidia-gfxG03-kmp-desktop, nvidia-gfxG03-kmp-default, nvidia-gfxG03-kmp-trace, and their Unified Memory variants. When installing cuda, the correct driver packages should also be specified. Without doing this, zypper will select packages that may not work on the system. Run the following to detect the flavor of kernel and install cuda with the appropriate driver packages:

   $ uname -r
     3.4.6-2.10-<flavor>
   $ sudo zypper install cuda nvidia-gfxG03-kmp-<flavor> \
                         nvidia-uvm-gfxG03-kmp-<flavor>

5. Add the user to the video group

   $ sudo usermod -a -G video <username>

6. Perform the post-installation actions.

1. Perform the pre-installation actions.
2. Enable armhf foreign architecture, if necessary

   The armhf foreign architecture must be enabled in order to install the cross-armhf toolkit. To enable armhf as a foreign architecture, the following commands must be executed:

   On Ubuntu 12.04,

   $ sudo sh -c \
     'echo "foreign-architecture armhf" >> /etc/dpkg/dpkg.cfg.d/multiarch'
   $ sudo apt-get update

   On Ubuntu 12.10 or greater,

   $ sudo dpkg --add-architecture armhf
   $ sudo apt-get update

3. Install repository meta-data

   When using a proxy server with aptitude, ensure that wget is set up to use the same proxy settings before installing the cuda-repo package.

   $ sudo dpkg -i cuda-repo-<distro>_<version>_<architecture>.deb

4. Update the Apt repository cache

   $ sudo apt-get update

5. Install CUDA

   $ sudo apt-get install cuda

6. Add libcuda.so symbolic link, if necessary

   The libcuda.so library is installed in the /usr/lib/nvidia-current directory. For pre-existing projects which use libcuda.so, it may be useful to add a symbolic link from libcuda.so in the /usr/lib directory.

7. Perform the post-installation actions.

Below are some additional capabilities of the package manager that users can take advantage of.

Before the stand-alone installation can be run, perform the pre-installation actions.

If you have already installed a standalone CUDA driver and desire to keep using it, you need to make sure it meets the minimum version requirement for the toolkit. This requirement can be found in the CUDA Toolkit release notes. With many distributions, the driver version number can be found in the graphical interface menus under Applications > System Tools > NVIDIA X Server Settings.. On the command line, the driver version number can be found by running /usr/bin/nvidia-settings.

The package manager installations (RPM/DEB packages) and the stand-alone installer installations (.run file) are incompatible. See below about how to uninstall any previous RPM/DEB installation.

The standalone installer can install any combination of the NVIDIA Driver (that includes the CUDA Driver), the CUDA Toolkit, or the CUDA Samples. If needed, each individual installer can be extracted by using the -extract=/absolute/path/to/extract/location/. The extraction path must be an absolute path.

The CUDA Toolkit installation includes a read-only copy of the CUDA Samples. The read-only copy is used to create a writable copy of the CUDA Samples at some other location at any point in time. To create this writable copy, use the cuda-install-samples-6.0.sh script provided with the toolkit. It is equivalent to installing the CUDA Samples with the standalone installer.

Exit the GUI if you are in a GUI environment by pressing Ctrl-Alt-Backspace. Some distributions require you to press this sequence twice in a row; others have disabled it altogether in favor of a command such as sudo service ligthdm stop. Still others require changing the system runlevel using a command such as /sbin/init 3 Consult your distribution's documentation to find out how to properly exit the GUI. This step is only required in the event that you want to install the NVIDIA Display Driver included in the standalone installer.

If you want to install the NVIDIA Display Driver included in the standalone installer, any previous driver installed through RPM or DEB packages MUST be uninstalled first. Such installation may be part of the default installation of your Linux distribution. Or it could have been installed as part of the package installation described in the previous section. To uninstall a DEB package, use sudo apt-get --purge remove package_name or equivalent. To uninstall a RPM package, use sudo yum remove package_name or equivalent.

To install any combination of the driver, toolkit, and the samples, simply execute the .run script. The installation of the driver requires the script to be run with root privileges. Depending on the target location, the toolkit and samples installations may also require root privileges.

By default, the toolkit and the samples will install under /usr/local/cuda-6.0 and $(HOME)/NVIDIA_CUDA-6.0_Samples, respectively. In addition, a symbolic link is created from /usr/local/cuda to /usr/local/cuda-6.0. The symbolic link is created in order for existing projects to automatically make use of the newly installed CUDA Toolkit.

If the target system includes both an integrated GPU (iGPU) and a discrete GPU (dGPU), the --no-opengl-libs option must be used. Otherwise, the openGL library used by the graphics driver of the iGPU will be overwritten and the GUI will not work. In addition, the xorg.conf update at the end of the installation must be declined.

If you wish to build all the samples, including those with graphical rather than command-line interfaces, additional system libraries or headers may be required. While every Linux distribution is slightly different with respect to package names and package installation procedures, the libraries and headers most likely to be necessary are OpenGL (e.g., Mesa), GLU, GLUT, and X11 (including Xi, Xmu, and GLX).

sudo apt-get install freeglut3-dev build-essential libx11-dev libxmu-dev libxi-dev libgl1-mesa-glx libglu1-mesa libglu1-mesa-dev

Check that the device files/dev/nvidia* exist and have the correct (0666) file permissions. These files are used by the CUDA Driver to communicate with the kernel-mode portion of the NVIDIA Driver. Applications that use the NVIDIA driver, such as a CUDA application or the X server (if any), will normally automatically create these files if they are missing using the setuidnvidia-modprobe tool that is bundled with the NVIDIA Driver. Some systems disallow setuid binaries, however, so if these files do not exist, you can create them manually either by running the command nvidia-smi as root at boot time or by using a startup script such as the one below:

#!/bin/bash

/sbin/modprobe nvidia

if [ "$?" -eq 0 ]; then
  # Count the number of NVIDIA controllers found.
  NVDEVS=`lspci | grep -i NVIDIA`
  N3D=`echo "$NVDEVS" | grep "3D controller" | wc -l`
  NVGA=`echo "$NVDEVS" | grep "VGA compatible controller" | wc -l`

  N=`expr $N3D + $NVGA - 1`
  for i in `seq 0 $N`; do
    mknod -m 666 /dev/nvidia$i c 195 $i
  done

  mknod -m 666 /dev/nvidiactl c 195 255

else
  exit 1
fi

Restart the GUI environment using the command startx, init 5, sudo service lightdm start, or the equivalent command on your system.

Once the stand-alone installation is complete, be sure to perform the post-installation actions.

The PATH variable needs to include /usr/local/cuda-6.0/bin

The LD_LIBRARY_PATH variable needs to contain /usr/local/cuda-6.0/lib on a 32-bit system, and /usr/local/cuda-6.0/lib64 on a 64-bit system

• To change the environment variables for 32-bit operating systems:

  $ export PATH=/usr/local/cuda-6.0/bin:$PATH
  $ export LD_LIBRARY_PATH=/usr/local/cuda-6.0/lib:$LD_LIBRARY_PATH

• To change the environment variables for 64-bit operating systems:

  $ export PATH=/usr/local/cuda-6.0/bin:$PATH
  $ export LD_LIBRARY_PATH=/usr/local/cuda-6.0/lib64:$LD_LIBRARY_PATH

Before continuing, it is important to verify that the CUDA toolkit can find and communicate correctly with the CUDA-capable hardware. To do this, you need to compile and run some of the included sample programs.