#ifndef _C_KERNEL_H_
#define _C_KERNEL_H_

        #define num_loops 1024
        #define MATRIX_COLS 1000
        #define num_threads_per_block 20

        __device__ float ReturnOut();
        __global__ void CKernel(float *c, float *c2, float *o)
        {

            int tx = blockIdx.x * num_threads_per_block + threadIdx.x;
            int ty = blockIdx.y * num_threads_per_block + threadIdx.y;

            //o[tx * MATRIX_COLS + ty] = ReturnOut();
            o[tx * MATRIX_COLS + ty] = 1.1;
        }

        __device__ float ReturnOut() {

            float out = 0.0;
            for (int ii=0; ii<num_loops; ++ii) {
                out += 1.5;
            }
            return out;
        }

#endif