ArrayFire

ArrayFire is a fast software library for GPU computing with an easy-to-use API. Its array-based function set makes GPU programming simple. ArrayFire is available for C, C++, Fortran, and Python and integrates with AMD, Intel, and NVIDIA hardware. Best of all, ArrayFire is free for most users!

ArrayFire is simple. A few lines of code in ArrayFire can replace dozens of lines of raw GPU code, saving you valuable time and lowering development costs. No need to waste time with building block functions. Plug-in ArrayFire and jump start your code!

Reasons to Use

ArrayFire for CUDA

Documentation: C/C++, Fortran, Python

Today, ArrayFire runs best with CUDA-capable NVIDIA GPUs. ArrayFire is designed to integrate easily with other CUDA code, but it can also be used without any other custom CUDA code. CUDA is the most advanced GPU computing platform and ArrayFire benefits from all the advance CUDA features.

ArrayFire for OpenCL

Documentation: C++

ArrayFire is also available, in beta form, for AMD and Intel hardware using OpenCL. Fewer functions are supported in OpenCL than are supported in CUDA, so please check the OpenCL documentation for supported feature list.

Details

• ArrayFire Free vs ArrayFire Pro.  What are the differences? Can I get by with the free version? See licensing page.
• Help Me.  Can I hire AccelerEyes engineers to help me? Check out our 1-day Guarantee program.
• HPC World-class Problems.  Scaling up to the world's largest supercomputers. See how ArrayFire for HPC rocks large problems.