Well let's take a look at the dimensions of this table.
Every node participates in every step, so each step takes order of N work,
and we're doubling the hop count on each step, thus there's log of N steps,
so it takes N log N work to construct the entire table.
Note this is more expensive than the serial algorithm which takes linear work, order of N work.
However, the serial algorithm also takes N steps,
whereas, we're finishing N log N steps here.