I'm a bit confused to read that because whilst they are hashed for reuse (requires optimising for ad hoc workloads to do this?), unless the text of the query is

I'm a bit confused to read that because whilst they are hashed for reuse (requires optimising for ad hoc workloads to do this?), unless the text of the query is an identical match it will need to compile another separate query plan and won't reuse the existing plan. This means you can end up in a situation where you have many ad hoc requests causing significant memory consumption. Would you be able to speak as to why prepared statements aren't necessary in MSSQL - I'm happy to link to some blog articles that are arguing the total other direction.