Introduction
In Microsoft Power BI, the semantic model underpins report and dashboard development. To produce the right reports and dashboards, you should have a clear understanding of the questions that your report and dashboard users will ask. Your next focus should be on the semantic model design that will best support the visualizations in your reports and dashboards. After you've arrived at a model design, you're ready to use Power BI Desktop to develop the model.
When developing the model, you will complete the following tasks:
- Connect to data
- Transform and prepare data
- Define business logic by adding Data Analysis Expressions (DAX) calculations
- Publish the model to Power BI
Understanding the structure of semantic models can help you design the right model to support your reports and dashboards. A semantic model can be developed in many ways, yet one or several of those ways are more optimal. Optimal models are important for delivering good query performance and for minimizing data refresh times and the use of service resources, including memory and CPU. The fewer resources that are used, the more models that can be hosted and at lower cost.