You should take a look at ParadisEO (http://paradiseo.gforge.inria.fr/) which is a highly scalable framework that incorporates numbers EAs, and supports various parallel architectures.
Thank you Mohamed-Mourad Lafifi, Drazen Bajer, and Henning Meyerhenke
Regarding to the Henning's comments, I would like to explain some points:
EAs techniques are not for simple optimization, they also are used for large-scale optimization, many-objective optimization, and natural language processing, and many other complex problems. Furthermore, recent development in EAs to optimize real word applications/problems indicate a lot of challenges to handle these problems with huge amount of data (data intensive applications). To solve such problems we need scalable algorithms based on Hadoop, Spark, and other technologies to get solutions.