1
edit
Changes
→Description
Another approach would have been to parallelise the whole nodes instead of the operations in them, but this is a little bit difficult because for each node to exist, the parent needs to have been evaluated already.
Learn from me and think long and hard about what you want to parallelise in regards of granularity, because in this case parallelising the individual operations in the Node made my algorithm significantly worse. It took a lot of programmer effort and gave me no reward in terms of speed. I did learn a lot about C++ writing this algorithm, so all was not lost.
==Source==