This project is under development as a graduation thesis at the Federal University of Ouro Preto.
The goal of the project is to develop methodologies and algorithms that accelerate OLAP queries over massive datasets. To achieve this, we use multiple clustered machines equipped with GPU devices to provide such speed-up.
The underling system used is the middleware JCL: used to distribute the database and operate the devices over a single interface on the cluster.
More information as well as documentation, would be available by the end of the year with a paper publication.