Am 27.12.2013 19:25, schrieb Daniel Kozak:
using OrderedAA improve speed 3x https://github.com/Kozzi11/Trash/tree/master/util
A possible downside of this implementation is though, that due to the fact that you are using a double linked list per index, there will be more chache misses during a read operation compared to a linear probing hashmap. Did you try using a array per index instead of a linked list, and measure if that makes any difference?
Kind Regards Benjamin Thaut
