Evaluation of Python-based tools for distributed computing on the Raspberry Pi
Jacob Morra, Qusay H. Mahmoud · Computer Science and Software Engineering · 2016
This paper presents the results of comparing the efficiency of four Python-based modules and libraries with regards to their respective ability to run distributed tasks across a given cluster of Raspberry Pi's. The four Python-based tools considered are the following: PyRO v4.45, DCM v1.0.0, PP v1.6.4.4, and Mpi4py v2.0.0. Three tests are used to compare run time efficiency for distributed tasks: The first test proposed is prime factorization of composite values n, where n is the product of two prime numbers, and p and q are the respective prime factors. The second test proposed is the approximation of π to a desired point of precision in a set number of iterations. The third test proposed is the approximation of π in a fixed number of iterations with variation in the cluster size used. The overall purpose of these tests is to provide a basis of comparison in regards to the efficiency of each Python-based tool.