For those in the know, the Octree structure (which Pointerra...

For those in the know, the Octree structure (which Pointerra use), whilst quite good, has significant limitations.

Octrees are certainly not the fastest tree structure around. The K-D Tree and the BVH Tree (Bounding Volume Hierarchy) are significantly faster to trace through with the ray casting algorithms.

Add to this the fact that the Octree structure (which Pointerra use) is much slower to rebuild. Hence the concept 100 billion points for or 400 billions or a billion billion points, is irrelevant. The structure is slower than competing trees and eventually impossible to update in any reasonable timeframe.

I suggest that Pointerra start redesigning their software to build a better tree, compression or no compression it makes no difference, and secondly they change their strategy. For the sake of the shareholders a different algorithm and a different approach is needed.

Just a suggestion to the uneducated.