add_vectors.cu

#include <stdlib.h>
#include <stdio.h>
#include <cuda.h>

#define N 512

// compile with nvcc add_vectors.cu

// forward declaration of CUDA kernel
__global__ void add_vectors(int* A, int* B, int* C);

int main(void) {
    // allocate space on CPU
    int* A = (int*) malloc(N * sizeof(int));
    int* B = (int*) malloc(N * sizeof(int));
    int* C = (int*) malloc(N * sizeof(int));

    // allocate space on GPU
    int* A_d;
    int* B_d;
    int* C_d;
    cudaMalloc(&A_d, N * sizeof(int));
    cudaMalloc(&B_d, N * sizeof(int));
    cudaMalloc(&C_d, N * sizeof(int));

    // populate data on CPU
    for (int i = 0; i < N; i++) {
        A[i] = i;
        B[i] = 2*i;
        C[i] = 0;
    }

    // copy CPU to GPU
    cudaMemcpy(A_d, A, N * sizeof(int), cudaMemcpyHostToDevice);
    cudaMemcpy(B_d, B, N * sizeof(int), cudaMemcpyHostToDevice);
    cudaMemcpy(C_d, C, N * sizeof(int), cudaMemcpyHostToDevice);

    // invoke kernel
    add_vectors<<<1, N>>>(A_d, B_d, C_d);

    // copy result from GPU to CPU
    cudaMemcpy(C, C_d, N * sizeof(int), cudaMemcpyDeviceToHost);

    // print result
    for (int i = 0; i < N; i++) {
        printf("%d\n", C[i]);
    }
}

__global__ void add_vectors(int* A, int* B, int* C) {
    C[threadIdx.x] = A[threadIdx.x] + B[threadIdx.x];
}