C++17 came with one really major extension for parallelism: execution policies for standard algorithms. Sixty nine algorithms were extended to accept execution policies in order to run parallel on multiple cores, and even with enabled vectorization.
For the user, this means that if we already use STL algorithms everywhere, we get a nice parallelization bonus for free. We can easily give our applications subsequent parallelization by simply adding a single execution policy argument to our existing STL algorithm calls.
In this recipe, we will implement a simple program (with a not too serious use case scenario) that lines up multiple STL algorithm calls. While using these, we will see how easy it is to use C++17 execution policies in order to let them run multithreaded. In the last subsections of this section, we will...
