If you are taking an analysis driven approach to algorithm tuning, there is a fantastic talk by Andrei Alexandrescu about the importance of caches coherency. He proposed a new formula for analyzing algorithmic performance. youtu.be/FJJTYQYB1JQ.
It's a C++ talk, but the lesson can be applied to any language.
Moral of the story: more work on a closer set of data is often much fast than less work spread across memory.
We're a place where coders share, stay up-to-date and grow their careers.
We strive for transparency and don't collect excess data.