Post Your Answer

Hello GPU programs are significantly different from the CPU programs. In short, the programs which run on CPU can’t be executed on GPU. And most likely, you can’t just run by MATLAB code either on the GPU. Also, the GPU code is written differently like CUDA or OpenCL. With the use of special functions from the parallel toolbox (which are not included in the MATLAB versions) can automatically executed on the GPU, only if your algorithms are based upon the functions. And one more point, CUDA only supports the nVidia graphic cards as said, and I can’t give a clear statement because I haven’t read up on this till now. Hope, the information will help you.

Hi  Sure you can try to use the parallel computing toolbox, if you developed the JPEG image compression using MATLAB. Using the toolbox, you can/might combine the calculations based on the CPU & GPU processors. You have to clear about the functions that, more and more functions for used in the CPU, might run on the graphics card. Indeed, you need to have a card that’s supported by a toolbox: following the requirements_ 1. Minimum of 1 GB RAM per MATLAB worker 2. Minimum of 5 GB space of disk 3. Maximum of 1 MATLAB worker per CPU 4. Latest CUDA driver You will also need to change the code for your program. Hope, you do your research as well based upon these points. Good luck