In the world of data warehouse and business intelligence terminology, there is quite a bit of ambiguity. When our group produces BI and business performance management solutions based on the complete Microsoft BI stack, we like to stick with solve particular business problems unique to a vertical industry. So we call the supporting database for the analytics, KPIs and BI a data mart as opposed to a data warehouse.
Our group is the Industry Solutions Group and the particular solution I am refering to is the Microsoft BI solution for vertical industries. The Microsoft BI stack is split-up into back-end database, ETL and analysis services (SQL Server product stack) and the front-end user interaction and analysis work through portals, which we call executive dashboards (Sharepoint, Proclarity and BSM, soon to be PerformancePoint Server).
Have a look on Microsoft.com for PerformancePoint Server which is currently in CTP 2. If you are interested in how such end-to-end Microsoft BI solutions really work and how to build a supporting data mart for such a solution, start with PPS. It will change the way that you look at business intelligence.
It is so easy to use as a tool that it seems impossible. And when you think about solving particular, easy to consume problems which BI can help solve, think data mart.