Topic Status: Some information in this topic is preview and subject to change in future releases. Preview information describes new features or changes to existing features in Microsoft SQL Server 2016 Community Technology Preview 2 (CTP2).
Tabular models are in-memory databases in Analysis Services. Using state-of-the-art compression algorithms and multi-threaded query processor, the xVelocity in-memory analytics engine (VertiPaq) delivers fast access to tabular model objects and data by reporting client applications such as Microsoft Excel and Microsoft Power View.
Tabular models support data access through two modes: Cached mode and DirectQuery mode. In cached mode, you can integrate data from multiple sources including relational databases, data feeds, and flat text files. In DirectQuery mode, you can bypass the in-memory model, allowing client applications to query data directly at the (SQL Server relational) source.
Tabular models are authored in SQL Server Data Tools (SSDT) using new tabular model project templates. You can import data from multiple sources, and then enrich the model by adding relationships, calculated columns, measures, KPIs, and hierarchies. Models can then be deployed to an instance of Analysis Services where client reporting applications can connect to them. Deployed models can be managed in SQL Server Management Studio just like multidimensional models. They can also be partitioned for optimized processing and secured to the row-level by using role based security.