Time difference?

Question

CUDA

Hello, can you please tell me what is the difference between:

clock_t start1 = clock();
...
clock_t end1 = clock();
  double seconds1 = (double)(end1 - start1) / CLOCKS_PER_SEC;
  cout << "Time:  " << seconds1 << " sec" << endl;

and

cudaEvent_t start, stop;
  cudaEventCreate(&start);
  cudaEventCreate(&stop);
  cudaEventRecord(start, 0);
...
        cudaThreadSynchronize();
  cudaEventRecord(stop, 0);
  cudaEventSynchronize(stop);
  float elapsedtime;
  cudaEventElapsedTime(&elapsedtime, start, stop);
  printf("Time = %3.1f ms\n", elapsedtime);
  cudaEventDestroy(start);
  cudaEventDestroy(stop);

I understand that one is for cuda, and the second is for the CPU, but what is their fundamental difference. Why do they measure the same piece of code differently (do they show different times?).
Thanks in advance for your reply, and sorry newbie for such stupid questions...

Answer 1

The fundamental difference is that the cuda interface is asynchronous, i.e. the time between start/stop will change the rate at which code is sent to the GPU.