I read it while looking up Judy Arrays. Their documentation is a little dated (circa 2000) but quite detailed. They mention specifically that 256-ary tree is very good in terms of cache-locality. Judy arrays are recommended to be used as sparse hash-tables though, so the use case is different from the one in your article.
What you said about cache-friendliness is true. I was thinking in terms of iteration.
Thank you for the link. I also found the post Hashtables vs Judy Arrays, Round 1 and it shows different scenarios in which either data structure excels.
It is safe to say that optimizing algorithms and data structures for the memory hierarchy, esp. CPU caches, is something where you can get a lot of performance boost. When I read Cache-Oblivious Algorithms and Data Structures I realized that there is a whole area of research around that and I know almost nothing :D
The link you shared is great. That will keep me busy for a while, thanks.
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.