Get certified for free when you join Fabric Data Days 2026 and dive into Fabric, Power BI, SQL, AI, and other essential data skills.
Join nowTry your skills in the Power BI Dataviz World Championship! Round one ends June 26. Join now
Looking for some insight in the Enhanced Comput Engine's CPU usage on a Capacity.
I (think) I'm finding that if I try to link to a large Dataflow (30M+ records, 20 columns), the Service is using the Enhanced Compute Engine and loading that table into memory. It pushes my P1 Capacity CPU to increase 10-15% and stay there.
Is that expected behavior that the CPU is consumed constantly by the Ehanced Compute Engine?
I've experienced this before where I was convinced linking Dataflows was the right idea until it throttled my Capacity.
| User | Count |
|---|---|
| 16 | |
| 14 | |
| 12 | |
| 9 | |
| 7 |
| User | Count |
|---|---|
| 39 | |
| 38 | |
| 34 | |
| 28 | |
| 21 |