cuda blockIdx.x 和 threadIdx.x

​​git​​

#include <stdio.h>

__global__ void hello_from_gpu()
{
    const int bid = blockIdx.x;
    const int tid = threadIdx.x;
    printf("Hello World from block %d and thread %d!\n", bid, tid);
}

int main(void)
{
    hello_from_gpu<<<2, 4>>>();
    cudaDeviceSynchronize();
    return 0;
}

nvcc hello4.cu
 ./a.out

$\begin{matrix} 00&01&02&03\\ 10&11&12&13\\ \end{matrix}$

Hello World from block 0 and thread 0!
Hello World from block 0 and thread 1!
Hello World from block 0 and thread 2!
Hello World from block 0 and thread 3!
Hello World from block 1 and thread 0!
Hello World from block 1 and thread 1!
Hello World from block 1 and thread 2!
Hello World from block 1 and thread 3!