Re:A cache miss is not a cache miss
by netheril96 on Sunday May 01, 2016 @11:50AM
(#11606616)
Is the comparison really fair? On a 64-bit system, the `std::unique_ptr<int>` takes 8 bytes, while a linked list node takes at least 24 bytes (as it must have a previous and next pointer). The larger sizes means less number of values can be in the cache at any time and may be the real reason for the slowdown.
[ Parent
| on HN ]
Re:A cache miss is not a cache miss
by larshagencpp on Sunday May 01, 2016 @12:52PM
(#11606864)
This is the reason that the problem was simplified to using vector<pair<int,int>> for all the experiments, so no difference in data size.