Optimizing data in the model

We've covered, in previous chapters, how Power BI supports directly connecting to a data store, such as connecting to an Enterprise Data Warehouse (EDW) that has billions of records and years of sales history. When configured with DirectQuery performance, Power BI becomes dependent upon not only the calculations and rendering of visuals in the report but also the performance of the underlying data store. If the EDW takes a long time to query, then the Power BI report that uses the query will also take a long time.

To help, Power BI also supports importing data, which reduces the performance dependency on the underlying data store to increase the performance of report visuals. Data imported into the Power BI data model is stored both in memory and on disk using the VertiPaq Storage Engine (SE). VertiPaq will compress data, sometimes by as much as 10x, so even with limitations on the amount of data imported, it's often able to store large datasets...