Today Python has become the predominant programming language to coordinate scientific applications, especially machine learning and deep learning applications. However, existing parallel programming paradigms, such as the Message Passing Interface (MPI), have proven useful for enhancing complex data flows when executed on multiple computers, including supercomputers. The framework is well known in the C-language community. However, many practitioners do not have the time to learn C to utilize such advanced cyberinfrastructure. Hence, it is advantageous to access MPI from Python. We showcase how you can easily deploy and use MPI from Python.
Laszewski et al. (Sat,) studied this question.