Forum Discussion
insert pivot table from power bi
You need to distinguish between a live connection to the semantic model and a DAX query against the model.
In the first scenario you treat the semantic model as a sort of OLAP cube, accessing it via MDX. That has a performance impact (as the MDX needs to be translated to DAX) and it can also lead to excessive resource use when your users discover the CUBEMEMBER functions and potentially use them in their Excel formulas.
The second scenario is much safer and you can create as many pivots from the DAX query as you want without materially impacting the semantic model (only impact is when you refresh the query).
Your usage pattern depends on how granular you need to go, how close to real time, and what performance expectations you have.
You need to distinguish between a live connection to the semantic model and a DAX query against the model.
In the first scenario you treat the semantic model as a sort of OLAP cube, accessing it via MDX. That has a performance impact (as the MDX needs to be translated to DAX) and it can also lead to excessive resource use when your users discover the CUBEMEMBER functions and potentially use them in their Excel formulas.
The second scenario is much safer and you can create as many pivots from the DAX query as you want without materially impacting the semantic model (only impact is when you refresh the query).
Your usage pattern depends on how granular you need to go, how close to real time, and what performance expectations you have.
Hello and thank you,
Definitely some opportunity cost in either one.
Zoey
