Setting up an ArrayFire project on Linux

Brian Kloppenborg ArrayFire Leave a Comment

Continuing our blog series "Learning ArrayFire from scratch" today we will discuss how to set up an ArrayFire project on Linux using CMake or standard makefiles.

Once you have installed ArrayFire on your system, the next thing to do is set up your build system. On Linux, you can create ArrayFire projects using almost any editor, compiler, or build system. The only requirements are that you include the ArrayFire header directories and link with the ArrayFire library you intend to use.

The big picture

On Linux, we suggest you install ArrayFire to the /usr/local/  directory so that all of the include files and libraries are part of your standard path. The installer will populate files in the following sub-directories:

Because ArrayFire follows standard installation practices, you can use basically any build system to create and compile projects that use ArrayFire. Among the many possible build systems on Linux we suggest using ArrayFire with either CMake or Makefiles with CMake being our preferred build system.

Prerequisite software

To build ArrayFire projects you will need a compiler

Fedora, Centos and Redhat

Install EPEL repo (not required for Fedora)

Install build dependencies

Debian and Ubuntu

Install common dependencies


We recommend that the CMake build system be used to create ArrayFire projects.f you are writing a new ArrayFire project in C/C++ from scratch, we suggest you grab a copy of our CMake Project Example; however, it is useful to read the documentation below in case you need to add ArrayFire to an existing project.

As discussed above, ArrayFire ships with a series of CMake scripts to make finding and using our library easy. The scripts will automatically find all versions of the ArrayFire library and pick the most powerful of the installed backends (typically CUDA).

First create a file called CMakeLists.txt  in your project directory:

and populate it with the following code:

where my_executable  is the name of the executable you wish to create. See the CMake documentation for more information on how to use CMake. Clearly the above code snippet precludes the use of both CUDA and OpenCL, see the ArrayFire CMake Example for an example of how to build executables for both backends from the same CMake script.

If you wish to use a specific backend, the find script also defines these variables:

Therefore, if you wish to target a specific specific backend, simply replace ${ArrayFire_LIBRARIES}  with ${ArrayFire_CPU} , ${ArrayFire_OPENCL} , ${ArrayFire_CUDA} , or ${ArrayFire_Unified}  in the TARGET_LINK_LIBRARIES  command above.

Next we need to instruct CMake to create build instructions and then compile. We suggest using CMake's out-of-source build functionality to keep your build and source files cleanly separated. To do this:

NOTE: If you have installed ArrayFire to a non-standard location, CMake can still help you out. When you execute CMake specify the path to the ArrayFireConfig*  files that are found in the share/ArrayFire/cmake  subdirectory of the installation folder. For example, if ArrayFire were installed locally to /opt/ArrayFire  then you would modify the cmake command above to contain the following definition:

You can also specify this information in the ccmake  command-line interface.


Building ArrayFire projects with Makefiles is fairly similar to CMake except you must specify all paths and libraries manually. As with any make project, you need to specify the include path to the directory containing arrayfire.h file. This should be -I /usr/local/include  if you followed our installation instructions. Similarly, you will need to specify the path to the ArrayFire library using the -L option (e.g. -L/usr/local/lib ) followed by the specific ArrayFire library you wish to use using the -l option (for example -lafcpu , -lafopencl , -lafcuda , or -laf  for the CPU, OpenCL, CUDA, and unified backends respectively.

Here is a minimial example MakeFile which uses ArrayFire's CPU backend: