Vector manipulation

Accelerated Computing CUDA CUDA Programming and Performance
1

What is the best way of adding two uint4 vectors? Also, is it possible to create a vector from a scalar,
for example
uint4 temp = uint4(5);

or to right shift a vector by a scalar

uint4 temp = uint4(32,32,32,32);
temp = temp >> 1;
2

You can simply perform element-wise operations

3

Thanks, I just miss the more terse way of doing it in OpenCL. I’m surprised there aren’t overloads for these operations.