Analysis Services has been Microsoft's enterprise Online Analytical Processing (OLAP) BI engine for many years. The Analysis Services Tabular model including the DAX programming language, in-memory and columnar storage, and columnar compression was first introduced with Microsoft's entry to self-service business intelligence with Power Pivot for Excel 2010. Analysis Services Tabular, which now includes the Power Query M programming language and the optional DirectQuery storage mode described earlier in this book, is the default installation mode of SSAS 2017. Additionally, the Analysis Services Tabular engine is built into Excel, Power BI, and the Azure Analysis Services PaaS offering.
SSAS also continues to support multidimensional mode instances, which consist of row and disk-based storage, and Multidimensional Expressions (MDX) for business logic. Despite structural advantages of tabular mode, multidimensional mode remains popular as it was available long before tabular mode and since it included many additional features relative to the initial releases of tabular mode. Power BI reports can leverage SSAS multidimensional models as Live connection data sources just like SSAS Tabular models and Microsoft has been clear that multidimensional mode will not be deprecated. However, multidimensional mode instances are currently not available in Azure Analysis Services and many new features and capabilities introduced in the past two years have been exclusive to tabular models.
Analysis Services Tabular, either via SSAS or Azure Analysis Services (Azure AS) instances, is recommended over multidimensional mode for new enterprise BI models. Similar to Power BI Desktop over Excel for self-service modeling and reporting, this recommendation is better aligned with Microsoft's BI roadmap and thus entails access to new features and capabilities. Additionally, given the common underlying engine between Power BI and Analysis Services Tabular, experience with Analysis Services Tabular positions organizations to better manage the relationship between Self-Service BI and Corporate BI.