Nice paper! Very clear and useful!!
I do not see absolute time values in the experiments, just speedups. It would have been nice to show some absolute values too.
I've implemented the algorithms [here](https://github.com/maxdan94/ParallelUnionFind).
On a very large graphs with more than 10 billions of edges (the twitter graph [here](https://sites.google.com/view/limass/datasets)), I observe that the verification algorithm is faster and leads to better speedups than the lock algorithm.
Nice paper! Very clear and useful!!
I do not see absolute time values in the experiments, just speedups. It would have been nice to show some absolute values too.
I've implemented the algorithms [here](https://github.com/maxdan94/ParallelUnionFind).
On a very large graphs with more than 10 billions of edges (the twitter graph [here](https://sites.google.com/view/limass/datasets)), I observe that the verification algorithm is faster and leads to better speedups than the lock algorithm.
Comments: