I'm interested in getting better performance out of SQLite on machines with more RAM, using the Python `sqlite3` module from the standard library. (Tips for more CPUs would be useful too - I'm interested in understanding what the benefits of paying for larger EC2 instances are, in addition to maximizing performance against currently available memory.) I've seen [PRAGMA cache_size](https://www.sqlite.org/pragma.html#pragma_cache_size) in the documentation - is this the main lever I should be looking at or are there other options that might help here? My current plan is to set up a simple load test that exercises my Python + SQLite code across a number of different queries, then modify the cache_size and see what impact that has on requests/second and request durations. I'd thoroughly appreciate any advice on ways to scale SQLite up like this. Most of my workloads are read-only (I open the SQLite file with the `?mode=ro` or `?immutable=1` flags) so I'm mainly interested in improving performance of SELECTs against read-only databases.