What is column encoding and compression conceptually?

< 1 min read

Column encoding and compression means storing data in a compact, optimized format to save memory.
In Power BI, VertiPaq engine encodes columns based on their data type and cardinality.
Low-cardinality columns, like Country, compress very efficiently.
High-cardinality columns, like Customer ID, compress less and consume more memory.
Encoding reduces storage size and improves query speed because fewer data blocks are scanned.
It also speeds up aggregations and DAX calculations.
Choosing appropriate data types (like integers instead of text) helps compression.
Overall, good encoding and compression improve model performance and scalability.

Power BI Essentials