A Strategy for Automatic Performance Tuning of Stencil Computations on GPUs

Algorithm 2

Generic 3D OpenCL stencil kernel.

_ _kernel void stencil_computation (

global float (in) [Y_SIZE] [X_SIZE],

global float (out) [Y_SIZE] [X_SIZE])

int x = get_global_id (0);

int y = get_global_id ;

int z = get_global_id (2);

float temp0 = in [z+] [y+] [x+];

float temp1 = in [z+] [y+] [x+];

⋮

float tempN = in [z+] [y+] [x+];

out [z] [y] [x]= f();