I am developing a model with a very large dataset using Direct Query. The report contains many measures and they seem to be processing sequentially rather than in parallel. To process the report in a reasonable time, it seems that I will have to precalculate all the measures in the databased and then just use one measure to return them all.
Currently, 1 measure is around 1 second, 50 measures in 50 seconds. If PBI could process in parallel, then I would hope that it would process in less than 5 seconds.
a) Are there any plans for PBI to process direct queries in parallel?
b) Any other solutions than precalculating the measures in the database?