To client applications, a cube appears to be a single collection of OLAP information. However, cubes are not stored as single entities. Dimension data is stored separately from aggregation and fact data. A cube’s aggregation and fact data is stored in one or more partitions, each of which can be stored in one of three modes.
A variety of options enables you to structure an OLAP data storage strategy that is efficient and effective for your organization’s requirements. The effects of changes to the data warehouse data upon which cubes are constructed vary depending on the storage choices you select.