Euclideon has a long history in this space of making relatively scammy claims out of technology that may seem far out to the average person, but is actually fairly bog-standard graphics techniques know by professionals in the field that aren't used because of their crippling drawbacks in the applications that people care about.

For instance, you can read this thread: https://news.ycombinator.com/item?id=2837948 and bear in mind that 2213 days later, the skeptics in that thread have basically been proved out by future results. (Although you can edit out the people who misunderstand octtrees and argue about the storage requirements for them.)

While we can't know what tech they're using for 100% certain, and they may have put their own twists on it, the most likely underlying tech they are referring to is sparse octtrees: https://stackoverflow.com/questions/985893/what-are-sparse-v... They are a real data structure. It is fantastic for static content and has amazing performance characteristics. However, if you are a programmer who has ever done any tree-type stuff, even if you haven't done 3D graphics, imagine trying to "animate" a single point by moving it in that structure. The algorithm to do so is quite simple... it's just impossibly expensive because of the staggering amount of tree traversal you must do. And due to the way the octtree subdivides, you can't even win with a vague handwave at "caching", though motivating that with math would take me longer than I have here. It completely fails to scale to anything like what the world today considers 3D graphics if you want any motion at all.

Further, there were other telltale signs that that was the technique they were using, such as repeated cells in the rendered video that were clearly just repeated cells in the octtree (a trivial elaboration to understand once you understand the data structure). If you looked you could even see them being aligned on powers of two.

Last I knew Euclideon had decided to change their focus to scanning high-resolution 3D images of things like cathedrals and such. I thought that was actually a good idea, it's a perfect fit for what their tech can actually do. Perhaps doing real engineering work and producing a real product of real value wasn't proving as lucrative as fraud. Certainly it's more work. They weren't even trying with that press kit, as others observe the tech described is going to be literally incapable of producing things that look like those concept photos.

You'd think they'd at least change their name, at this point.

I'm also hoping the Euclideon Defense Squad is tired out at this point. (Well, it's not really the Euclideon Defense Squad. It's really the Squad of People Annoyed That Some People Claming Things Are Impossible, What If You're Wrong, They Thought We'd Never Break The Sound Barrier, Didn't They? It just happens to be that the Euclideon was the topic du jour. But that's a bit long. It also tended to miss the fact that even if what they're describing is possible that doesn't mean Euclideon are the people to do it....) They're scammers. It's pretty much proved at this point, I don't even feel like that's a libel risk. They've not produced what they said they will in the past, and if this is their current focus, they will continue to not produce it. Since what they were trying to produce in the past was impossible, their failure to produce it was no great surprise. This table they're describing may be possible, it just won't be as impressive as they are promising. Possibly useful, though. Unless they really were stupid enough to tie it to their voxel technology, in which case awhoop there goes most of the use cases again.