Limitations of ER modelling while modelling a dwh

Could anyone help me out of Limitations of ER modelling when used for modelling data in a dwh

There are none I can think of.

The downside is query performance. Those who subscribe to the Inmon approach to a data warehouse, that is, maintaing a somewhat 3NF data repository of all data, do not (or should not) allow end-user access to the repository for data analysis queries. There are two reasons for this: 1, the schema tends to be overly complex and difficult to use, and 2, massive queries run for a very long time. Inmon recommends that end-user accessible data be published in the form of data marts (usually dimensional, usually aggregate - star schema or cubes) or other data extracts for consumption.

Ahh, when someone says ER-modeling, that doesn't imply dimensional or normalized data models.