Binary Trees Implementations Comparison for Multicore Programming

Although there are a lot of optimized concurrent algorithms for lists, queues, stacks, hash tables and other common data structures, there are much fewer optimized concurrent Binary Search Trees implementations. This paper compares several concurrent implementations of Binary Search Tree. The implementations are also compared to a concurrent skip list, which is the general way to implement a concurrent set or map. All the data structures are implemented in C++ using Hazard Pointers to manage memory. The different implementations are compared using several measures to determine the best performing structures in different situations. The benchmark shows that some data Binary Search Tree implementations are performing much better than the concurrent skip list base implementation. The optimistic AVL Tree that has been tested is shown to be performing very well. Baptiste Wicht Computer Science Master degree Switzerland HES-SO University of applied science Baptiste Wicht 1 of 69 June 7, 2012