How to design data warehouses?

Question

SQL Server

Example:
There is a data warehouse that consists of one huge table (and there are measurements and facts).
And there is a normalized HD, where the facts are in one table, the measurements are spaced in many.
I think it's called the Star scheme?
The question is, in which of these cases will it be faster to work on an insert, and in which one on Select?
If anyone knows literature on this topic, please recommend.

Answer 1

Max, the insert in both cases will be in one table, the data is not inserted immediately into the heap - well, usually. If done, then the insert turns into a transaction. The answer to the question should be obvious by now, right?
And further. The insert slows down when there are indexes. Because these indices are recalculated.
To shove all the data into one table is to learn bad things. Especially if it's big.