A data warehouse can best be understood as a distribution center; this distribution center stands between the creation of data, and the various departments/teams within a business that find ways to leverage it. Understandably, there have been many variations of this model,
including: CIF (Corporate Information Factory), Dimensional Data Warehouses and Standalone
Marts. Interestingly, the existence of one architecture does not necessarily mean the absence of others. For example, CIF architectures may utilize Data Marts for boutique business requirements.
CIF is synonymous with corporate structures, in my opinion; Massive centralized data, accessible to all arms of the business with an emphasis on uniform access.
Dimensional Data Warehouses are unique, in that they also rely on a more visual modeling of data for structure. As a result, these structures can be very user-friendly. Specifically, Dimensional Data warehouses prioritize speeds of query responses.
"Standalone" Data Marts are very apty named, existing in more of a "plug and play" style of architecture. Data Marts can be implemented into other, more centralized structures. This type of data architecture is often automated, as they are usually designed with a specific purpose in mind, rather than a user/query relationship.
