Creating a process pool with multiprocessing

Multiprocessing is a standard Python module that targets machines with multiple processors. Multiprocessing works around the Global Interpreter Lock (GIL) by creating multiple processes.

Note

The GIL locks Python bytecode so that only one thread can access it.

Multiprocessing supports process pools, queues, and pipes. A process pool is a pool of system processes that can execute a function in parallel. Queues are data structures that are usually used to store tasks. Pipes connect different processes in such a way that the output of one process becomes the input of another.

Note

Windows doesn't have an os.fork() function, so we need to make sure that only imports and def blocks are defined outside the if __name__ == "__main__" block.

Create a pool and register a function as follows:

   p = mp.Pool(nprocs) 

The pool has a map() method that is the parallel equivalent of the Python map() function:

p.map(simulate, [i for i in xrange(10, 50...