[vjerancrnjak]

Filtering by user id would also be trivially fast.

It’s transactions mostly that make things slow. Like various isolation levels, failures if stale data was read in a transaction etc.

I understand the difference, just a shame there’s nothing close to read or write rate , even on an index structure that has a copy of the columns.

I’m aware that similar partitioning is available and that improves write and read rate but not to these magnitudes .

[FridgeSeal]

Some of the “new SQL” hybrid (HTAP, hybrid transaction-analytical processing) databases might be of interest to you. TiDB is the main example off the top of my head.

[beoberha]

look at who you’re arguing with ;)