There is a query cache. But I sorta doubt this benchmark took appropriate advantage of it. We don't know enough about how the author is running postgres to be sure. But the reported postgres numbers are pretty slow for machine-local SELECTs. Maybe Django's ORM is getting in the way?

> Why can't Postgres implement a transparent caching layer similar to Redis?

Bruh it does! That's what page pools are all about. Not to mention the query cache...

How would you cache a write?

You can't cache it, but if you don't care about durability as in this case you can acknowledge it before it is fsync'd.

A write is really fast if you don't do it.