Help with structs benchmarks

[al2o3cr]

Here’s some prior discussion that seems like a good starting point:

Big maps versus small maps performance

I tweeted this morning about some BEAM internals (https://twitter.com/akoutmos/status/1266034402422853633). Specifically about how maps are represented internally as hash array mapped tries (or HAMT for short) when the size of the map is greater than 32 entries. @StanBright had a good idea to run some benchmarks against small maps (32 entries) and large maps (greater than 32 keys). I threw together a quick benchee script (found here https://gist.github.com/akoutmos/6e965558fa8bb5771e90b1961762a…

A big change worth learning about for this specific case (Elixir structs) is the new_map_lit machinery that landed in OTP 21:

Introduce new_map_lit operation in the loader