Both projects are still pretty nice so plenty of optimization work left to be done.

Spark is starting to amass all the heavy hitters in big data so no doubt it will get much faster over time.