What my code does is reading a struct instance from the vector and editing the two integer fields in the sub-struct.
This does not sound like awful lot of computation. I think that single thread will handle the task as fast or perhaps faster than multiple threads...
Now if you had 3 millions of elements instead of 300...
(Of course, I can still be mistaken about the amount of computation per element performed. Have you benchmarked that (I mean, single-threaded time to do one element)?