Without seeing your schema, that sounds like a bit much, but maybe not. That is kind of the point of a relational DB.

I always attempt to normalize as high as possible first, and then denormalize IFF it’s necessary for performance. Postgres has a default join limit (i.e. where it shifts to the genetic algorithm for joins, and may not be optimal) of 8, but I’ve still seen it do fine far above that.