How to run multiprocessing Pool in one of the threads?

Question

Python

Is it possible to run a multiprocessing Pool in one of the threads and at that moment give it all the resources? The remaining threads are relatively rare I / O bound (appear as a result of array processing) tasks. It is necessary to achieve maximum performance for array analysis. To be honest, I don’t quite understand yet, since visually if you run multiprocessing in one of the threads, then all the others seem to be blocked automatically, since the execution thread completely hangs on the process pool (but this is not accurate).