Vertical partitioning involves splitting a table into multiple tables with fewer columns and using additional tables to store columns that relate rows across tables. We commonly refer to this as a join operation. We can then store these different tables in different nodes.

Normalization is one way to perform vertical partitioning. However, general vertical partitioning goes far beyond that: it splits a column, even when they are normalized. 

▪ Vertical partitioning is a "Row Splitting".

▪ Each partition holds a subset of the columns for table in the database.

▪ The data divided based on columns and we can divide by mostly visited columns and the other columns in different servers.

▪ Frequently accessed columns can be use in one vertical partition and less frequently accessed fields in another.

▪ Example case: Facebook, the user profile data can be in different partition and user mostly visited data can be different server.

▪ Benefit of Vertical Partitioning is you can separate the critical and mostly visited columns in a separate server.

▪ Best practice: Divide by rarely-changes and frequentlychange columns into different servers. 

https://blog.bytebytego.com/p/vertical-partitioning-vs-horizontal