Geek Answers Handbook

The easiest way to use a GPU for a parallel loop

I currently have a parallel loop like this:

int testValues[16]={5,2,2,10,4,4,2,100,5,2,4,3,29,4,1,52};
parallel_for (1, 100, 1, [&](int i){ 
    int var4;
    int values[16]={-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1};
    /* ...nested for loops */
        for (var4=0; var4<16; var4++) {
            if (values[var4] != testValues[var4]) break;
        }
    /* ...end nested loops */
}

I have optimized as much as I can, to such an extent that the only thing I can do is add more resources.

I am interested in using the GPU for parallel task processing. I read that embarrassing parallel tasks like this can efficiently use a modern GPU.

Using any language, what is the easiest way to use a GPU for a simple parallel for a loop like this?

I don't know anything about GPU architectures or GPU native code.

+5
source share
3 answers

Li-aung Yip , - Matlab, ( ) GPU. .

GPU CUDA OpenCL ( OpenCL AMD). , , , OpenCL ( C). /. OpenCL AMD ...

, . , , , , , gpu. , , , . , , - theano - . , python/numpy, , theano, ( "" - , , , , ).

ps , gpu , btw.

+2

, .

http://www.accelereyes.com/products/arrayfire

openCL, : Intel, AMD Nvidia.

+1

, OpenACC, parallelism . (C/++/Fortran) , , . . OpenACC - GPU, parallelism , , CUDA OpenCL. OpenACC PGI, Cray CAPS (PGI 30- ).

0

More articles:


All Articles