The large volumes of data that nowadays exist in transactional systems of companies increasingly require data warehouses with low response times, capable of handling large volumes of information, with the objective to give support to Business Intelligence tools that are designed for decision making. This thesis aims to compare the technologies that exist so far, commonly used by companies, and new emerging database technologies that allow analyze large volumes of data with high performance. Basically, the objective is to know in which circumstances is useful a technology or another, their advantages and disadvantages, the type of data they support, among others. This project has mainly two goals. The first goal is to compare four OLAP architectures in a theoretical way, in particular ROLAP, MOLAP, column-based architecture and associative model architecture. The second one is to choose a software system of each OLAP architecture above, install them in identical virtualized environments, perform a predefined set of tests with the same data source, and finally, study their behavior in order to validate the theoretical comparison. Once the results have been obtained, it is intended to provide a usage list of the OLAP architectures studied, both traditional and newer ones, specifying their most important features, in order to know in which cases it is more suitable to use each one. This project is developed during an internship at Everis.