A local instance of Analysis Services Tabular is used by Power BI Desktop when creating both import and DirectQuery data models. When these models are published to the Power BI Service as Power BI Desktop (.PBIX) files, the Power BI service extracts the Analysis Services database from the Power BI Desktop file and provides an Analysis Services server instance for running the database in Power BI. However, despite this very deep integration with Analysis Services, Power BI Desktop is primarily targeted at business analysts and self-service BI, while Analysis Services is intended for business intelligence professionals and enterprise BI solutions.
Given these different target personas and use cases, many features and capabilities of Analysis Services Tabular models are not available to Power BI Desktop. For example, BI/IT organizations generally utilize robust version control systems such as Visual Studio Team Services (VSTS) and multiple tools and programmatic interfaces to manage and administer their solutions. Additionally, there are modeling features of Analysis Services including Perspectives, Display Folders, and KPIs that are currently not accessible to dataset designers of Power BI datasets.
The following table compares Power BI Desktop models to Analysis Services (Tabular) models across 19 features:
As shown in the table, Analysis Services Tabular models, whether deployed to SSAS or Azure AS instances, are able to provide both maximum modeling functionality to aid the user experience as well as rich administrative tools and interfaces. It's outside the scope of this chapter to describe each feature, but the following three sections summarize the primary incremental benefits provided by Analysis Services over Power BI datasets built with Power BI Desktop.
In addition to the features identified in the table and described in the following sections, Analysis Services models avoid the dataset per app workspace limitation described in the Corporate Power BI datasets section earlier. A single Analysis Services model and its refresh process can be leveraged across multiple app workspaces and their corresponding Power BI apps.