that's a good idea. how might you test this? could also include a codex memory test.

I'm guessing having a portable memory that's comparable with first party memory is the goal.