Iteration Heatmap Recursion Experiment

Uses a brickmap (4x4x4) SVO, uses recursive DDA for hierarchical traversal. SVO buffers are generated on the CPU for now. Traversal speed is okay; roughly 3-6ms for this scene at 1440p on an AMD 780M. Spent a week trying to optimize this using different traversal methods and DDA always was the fastest one, albeit it has much lower theoretical occupancy than the others. This isn't a problem in of itself, but the amount of buffer reads we are doing definitely add up, so hiding that latency would be a nice thing.