cuda_main_call_kernel.cu

int main() {

	int n1, n2, sum;
	int *dN1, *dN2, *dSum;

	int size = sizeof(int);

	// allocate device memory
	cudaMalloc((void **)&dN1, size);
	cudaMalloc((void **)&dN2, size);
	cudaMalloc((void **)&dSum, size);

	n1 = 10;
	n2 = 15;

	//copy from host memory to device memory
	cudaMemcpy(dN1, &n1, size, cudaMemcpyHostToDevice);
	cudaMemcpy(dN2, &n2, size, cudaMemcpyHostToDevice);


	dim3 Dg(2, 2, 1);
	dim3 Db(2, 2, 4);

	getSum << <Dg, Db >> > (dN1, dN2, dSum);

}