Thank you for your reply.
Actually, I couldn't recognize the concrete maximum value of np.int32.
The fact that np.iinfo(np.int32).max / (27 * 450 * 450 * 450) is smaller than
1 tells me the issue come from Overflow.
Changing "np.int32" to "np.int64" and "int" to "long long" solves the bug.
I
takayanagi.tets...@jp.panasonic.com writes:
> Hi, All.
> I have developed Lattice Boltzmann Method Code with PyCUDA in our company for
> simulating Air flow.
> Then, I need to handle large gpuarray such like arr[velocity][Z][Y][X] for
> 3-dimensional fluid flow.
> My code run correctly