Optimising DAX fusion - Power BI
summary
Uses DAX Studio and server timings to show concrete steps that cut 337 storage-engine calls to one and drop server time from ~21s to about 1–2s on a heavy Matrix. Explains regular fusion vs horizontal fusion, fixes import/direct-query islands by setting dims to dual, replaces calculation groups with a disconnected measure table, and applies a sumx iterator pattern plus a small time‑intelligence table (mini many‑to‑many) to inject a tight WHERE clause so the workaround actually works.
AI-written summary · about
chapters (9)
- 0:00Introduction to fusion
- 2:17Defining regular fusion
- 6:21Horizontal fusion concepts
- 10:03Workaround for broken fusion
- 16:48Analyzing a slow report
- 21:22Optimizing data models
- 28:38Advanced iteration patterns
- 37:26Time intelligence optimization
- 43:26Final summary and questions
related talks
51mLessons learned from optimizing enterprise data models
19mDAX in 20 Minutes - The Hard Parts
51mPower BI Composite and Hybrid Models
50mThe Path to Success with Power BI: Best Practices
51mTroubleshooting Power BI Report Performance
45mAzure DW and Power BI, reporting at the speed of light!
50mTroubleshooting Power BI Reports
48mCombine Power BI with External Tools