Efficient C++ LRU cache. Gets usually don't cause any memory writes or internal cache restructuring (ie. no linked-list). Optimal workloads with 100% hit rate cause zero memory writes.
This is a header-only component; only flipfloplru.h is required.