nine_nine_game.cu

/*
  multiplication table using CUDA
  refer : http://blog.daum.net/heoly/7 (Thank you)
*/
#include <stdio.h>
#include <malloc.h>
#include <cuda_runtime.h>

#define BLOCK_SIZE 8
#define THREAD_SIZE 9

// Device code
__global__ void test(int *result)
{
 int tidx, bidx;
 tidx = threadIdx.x; //x-coordinate of thread
 bidx = blockIdx.x; //x-coordinate of block

 result[THREAD_SIZE * bidx + tidx] = (bidx + 2) * (tidx + 1);
}

// Host code
int main()
{
 int *host_Result; //Save result data of host
 int *device_Result; //Save result data of device

 int i=0, j=0;

 //Allocate host memory
 host_Result = (int *)malloc( BLOCK_SIZE * THREAD_SIZE * sizeof(int) );

 //Allocate device memory
 cudaMalloc( (void**) &device_Result, sizeof(int) * BLOCK_SIZE * THREAD_SIZE);
 
 //Function name <<BLOCK_SIZE, THREAD_SIZE>>> parameters
 test <<<BLOCK_SIZE, THREAD_SIZE>>>(device_Result); //Execute Device code
 
 //Copy device result to host result
 cudaMemcpy( host_Result, device_Result, sizeof(int) * BLOCK_SIZE * THREAD_SIZE, cudaMemcpyDeviceToHost );
 

 //Print result
 for(j=0; j<BLOCK_SIZE; j++)
 {
  printf("%3d step\n", (j + 2));
  for(i=0; i<THREAD_SIZE; i++)
  {
   printf("%3d X %3d = %3d\n", j+2, i+1, host_Result[j * THREAD_SIZE + i]);
  }
  printf("\n");
 }
 free(host_Result); //Free host memory
 cudaFree(device_Result); //Free device memory

 return 1;
}