d = vec_sub_u128(a, b)
Subtracts unsigned quadword values.
This function operates on the vectors as 128-bit unsigned integers. It returns low 128 bits of a - b.