On-Line Analytical Processing (OLAP) systems are database-like systems optimised for querying in aggregate. Normally OLAP systems use specialised query languages and APIs, different from relational database systems.
Several different architectures are used in OLAP systems:
MOLAP (Multi-dimensional OLAP) systems persist the data in a specialised data store, and calculate and persist aggregate rollups of the data that can be used for faster querying. Examples of MOLAP systems include Microsoft SSAS and Hyperion Essbase.
ROLAP (Relational OLAP) systems front a standard database system and issue queries to the underlying database. Examples of ROLAP systems include Business Objects and Microsoft Report Builder.
HOLAP (Hybrid OLAP) systems query data from an underlying database store but calculate and cache aggregates in memory for performance. Examples of HOLAP systems include Mondrian and Microsoft SSAS.
In-memory OLAP systems read data from sources and maintain an in-memory data structure for fast queries. Examples of in-memory OLAP systems include QlikView and Microsoft PowerPivot.