The performance of a Power BI semantic model – formerly known as dataset – depends on a variety of factors. One of the biggest influences is the data model itself: did you put everything into one single big table (you might found out writing DAX might get complicated), did you use a star schema as the foundation or did you just import all the tables in 3rd normal form from the source system (you might found out writing DAX might get really complicated).
I’ve been advocating to use always use the star schema unless your use case dictates otherwise. Using