What are the key differences between Power BI datasets and lakehouse tables in Fabric?

• A. Lakehouse tables are only for machine learning.
• B. They are the same.
• C. Datasets are unstructured data.
• D. Power BI datasets are optimized for BI analytics within Power BI; lakehouse tables are raw/curated data sources for broader analytics and governance accessible by multiple experiences.

Answer: (D)

The main idea is understanding how Power BI datasets and lakehouse tables serve different usage scenarios in Fabric: Power BI datasets are optimized for BI analytics inside Power BI, while lakehouse tables are raw or curated data sources designed for broader analytics and governance accessible by multiple experiences.

Power BI datasets provide the data model, relationships, and DAX-based measures that power interactive reports and dashboards, with an in-memory engine tuned for fast visuals and security features like row-level security. Lakehouse tables, on the other hand, live in the lakehouse and support governance, data quality, and access from various tools and experiences (such as notebooks, SQL engines, and BI tools), handling raw or curated data for broader analytic needs.

The other statements don’t fit because lakehouse tables aren’t restricted to machine learning, they aren’t the same as Power BI datasets, and datasets aren’t inherently unstructured—datasets are structured, optimized data models for BI analysis.