Performance issue with table visual

[Quiny_Harl]

Hi pbi folks,

I have the following snowflake part of my data model:

I have a table visual where I have DIM_PRODUCT columns and FACT 1 measures. I would like to add a few more measures from the PRODUCT_PACKAGING table. They are simple measures, for example SUM(PRODUCT_PACKAGING[Boxes]) and SUM(PRODUCT_PACKAGING[Frames]). However, when I try to add the measures to my table visual, I get the "there is not enough memory" error and I don't understand why. The table visual is already quite heavy (with many columns and measures), but it is working with all other measures, except these from the PRODUCT_PACKAGING table. 
I tried another approach where instead of having a seperate PRODUCT_PACKAGING dim table, I just left join it to DIM_PRODUCT in Power Query. Then I created, for example, a measure which is SUM(DIM_PRODUCT[Boxes]). The result is exactly the same - the table visual won't load when I add this measure to it.

[croberts21]

This is for PBI Desktop, right?

Measures are calculated on the fly and if you have too many records shown, you may run out of memory on a PC. I recommend memory of 32GB for a PC and PBI Desktop. I also recommend installing PBI on an SSD which is faster than a HDD. Column values are calculated and stored. 

I'm not sure if PBI calculates the measure for all records in the dataset, or just for the records in a visual. If it were only calculated for records in a viz, you could use the Filter pane to reduce the number of records shown, but when a user shows more records, they will likely run out of memory also.

We regularly get an "out of memory" error on the PBI Service when refreshing some datasets that use one of our large tables but that's because of the level of service we bought, we only have 4GB of memory. 

[Quiny_Harl]

- This is Power BI Desktop, but if I post the report to Premium Workspace in Power BI service, I still get the same error.
- Filtering the number of records is not an option in this case, although even when I filter down the number of records it still doesn't want to load.
- Yes, there are a lot of records shown, but the question is why I don't have this error with other measures but only with the measures from the PRODUCT_PACKAGING table.