Variant with bounding volume hierarchy such as a k-d tree?

[Qqwy]

I very much liked the talk. I would be really interested in how these three techniques would compare performance-wise, if a bounding volume hierarchy, like a k-d tree were to be introduced. My expectation is that all three versions will speed up dramatically; however I wonder which one of the three is able to benefit from it the most.

[mattgodbolt]

Absolutely: it was always my intention to add this after to see how easy it was to add to the three approaches. In the end I didn't have time, and I wanted to get to the bottom of the performance issues found in the DoD version.