Doing redundant work to speed up distributed queries
It turns out that repeating calculations can actually result in better performance when dealing with large scale distributed computing. Cool!
