Update maximum CUDA grid size
According to the documentation, the maximum CUDA grid size is
(2147483647, 65535, 65535). So the getMaxGridSize method should should be updated to report 2147483647 for the x dimension whenever possible.
For devices with 2.x compute capability, which are unsupported since CUDA 9.0, the maximum grid size was
(65535, 65535, 65535).