How to make sure async threads do not use shared resources and utilize all CPU cores?

Question

I am using std::async to parallelize calling a rendering function that takes in two integers (i.e. x and y of a pixel in an image) and