Errors on examples

Accelerated Computing CUDA CUDA Programming and Performance
1

Im having trouble with running my code. Im using the book called “CUDA By Example” and was doing several examples.
Among those examples, those of which run on CPU was working great except that when I used global to make it run on
GPU it starts to give wrong values. Following is one of the program which is supposed to add values from array a and b
and store the value to c.

Adding.cu (1.06 KB)

I have no idea what is going on. Did anyone had similar problem? Im using nvcc compiler. Can someone please help me?
Thank you.

2

Try to put
cudaThreadSynchronize();
after
add<<<N, 1>>>(dev_a, dev_b, dev_c);

3

I put cudaThreadSynchronize() after add<<<N,1>>>(dev_a, dev_b, dev_c) but still the values seems odd. Is there

any other solutions??

4

Your code is OK. This is output on my computer

0 + 0 = 0
-1 + 1 = 0
-2 + 4 = 2
-3 + 9 = 6
-4 + 16 = 12
-5 + 25 = 20
-6 + 36 = 30
-7 + 49 = 42
-8 + 64 = 56
-9 + 81 = 72