Forum Discussion
frithjof_v
2 years agoCommunity Champion
Power Query - Compare capacity units (CU) usage
Hi all, Let's say I need to do data transformations in Power Query for my new report. Let's say I am focusing specifically on one table in my semantic model, which requires some transformations...
lbendlin
2 years agoSuper User
The answer is a definitive maybe.
- Ideally place your comparative pbix into different workspaces
- Ideally run all of the tests at the same time
- repeat your tests at least ten times, throw away the outliers and average the rest
(yes, assuming you have Premium/Fabric capacity and are using the metric app)
Or - examine your Power Query code and optimize it manually. There are lots of things you can do like rearranging the order of the steps and liberally applying Table.Buffer and its siblings.
I still like the CU based approach more as it gives you a true cost number. It will be a combination of time spent and computational complexity.