Accelerating matlab with cuda

Publication Year: 


     MATLAB® is a powerful tool for prototyping and analysis. MATLAB could be easily extended via MEX files to take advantage of the computational power offered by the latest NVIDIA graphics processor unit (GPU). The graphic processor can be considered as a compute device that is capable of efficiently executing data parallel computations, and using CUDA, (Compute Unified Device Architecture) can be programmed using a C-like language. MEXfiles are a way to call custom C or FORTRAN code directly from MATLAB as if they were MATLAB built-in functions.
     This presentation will show the feasibility and benefits of using this approach: with the achieved speed-up, there is no need to recode the application in C or FORTRAN.


Nvidia, University of Utah