Cloud-based Evolutionary Parallel Computation Using Low Cost Storage Services

In this paper we will check several different, and difficult, problems, and present what effects the automatic load-balancing and asynchrony have on the speed of resolution of problems. We conclude that the effects of parallelization are noted with difficult problems, although there is a complicated interplay between the number of nodes and migration rate that makes difficult to say exactly what is the best combination of them. This paper uses Drop box as free storage service to conect several heterogeous computers to run an evolutionary algorithm. The algorithmic performance of the system has been proved in previous papers.