Slow loading report - Should I put more dependence on SQL Queries?

[HotChilli]

I thought there was only one table in the original post?

I am concerned about this statement "with a table visual that has 20 columns, each column is a DAX measure"  - they can't all be measures, can they?  Some fields must be from dimension tables.

How many Fact tables involved ?

How many rows in the Fact tables?

What about identifying which individual measures are slow by starting with a cut down version of the table and adding measures one-by-one?

How about posting a cut down sample file and I'll have a look?

[HotChilli]

And performance analyser results?

[Anonymous]

@HotChilli  DAX Query = 36815 for one table and the other table DAX Query = 115051

[HotChilli]

Maybe.  Examine your model to see if the Fact table(s) have the desired granularity , for example, if you never analyse the data at a daily level, summarise it at weekly/monthly level to reduce the number of rows.

Also, use performance analyser on the View menu to see what's taking up time and define 'forever'. Write these things down to act as a baseline.

Is the model a star schema? Are you using appropriate data types etc.? 

[Anonymous]

thanks for the response @HotChilli  the fact tables do have the corrent grainularity, it takes about 2 min. for the table to load & using the correct data types. Yep star schema