Summary

In this chapter, we learned how to speed up semantic models for Import mode. The theory from the Kimball Group was used frequently for dimensional modeling. The star schema provides an efficient use of data for semantic models. This format works well with the Analysis Services engine used in Power BI. The four-step dimensional modeling process from Kimball provides practical examples we can use for optimal performance.

Then, we focused on reducing the size of the semantic models. This is important because less data means less processing, which results in better performance and more free resources for other parallel operations. We also explored techniques to help Analysis Services compress data better, such as choosing appropriate data types, reducing cardinality for columns, and preferring numbers over text strings.

Lastly, we learned how to optimize RLS. We learned that RLS works just like regular filters and that previous guidance about fast relationships also applies...