Wolfram Mathematica More about Mathematica »
More User Stories »

CUDA & OpenCL Programming in Mathematica

Simplify Access to Performance Computing in Industry and Research

Mathematica combines programming ease-of-use with the computational speed of GPU-equipped hardware to dramatically increases application performance and user productivity.

Mathematica simplifies GPU programming with CUDALink and OpenCLLink—a high-level interface to the GPU built on top of Mathematica's language and development technologies. They allow you to execute code on your GPU with minimal effort.

By fully integrating and automating the GPU's capabilities using Mathematica, hiding unnecessary complexity associated with GPU programming, and lowering the learning curve for GPU computation, you experience a more productive and efficient development cycle.

Computers that have CUDA or OpenCL support are supported by CUDALink or OpenCLLink with Mathematica. CUDALink and OpenCLLink are supported on Windows (32- and 64-bit), Mac OS X (64-bit), and Linux (32- and 64-bit).

Download our free CUDA or OpenCL white papers to learn more about Mathematica's GPU computing capabilities.


Mathematica GPU programming features include:

  • Intuitive CUDALink and OpenCLLink programming interface
  • Automatic memory management, multicore computing, and built-in functions for many applications
  • Elimination of the need to write C/C++, FORTRAN, or CUDA/OpenCL wrapper code for GPU computing
  • Support for both single and double precision
  • Integration with Mathematica without the need for add-ons