http://www.youtube.com/watch?v=UKUuUvDSXk4

Saw this video on the weekend about this new Australian company and what they see for the future of computer graphics. They don't get too technical on how they are doing what they are doing but it seems pretty neat. I just don't get how they can talk about how many "atoms" make up an object (which is a crazy high number) but then it seems to run fairly fast, where as if you did it with polygons all the sudden computers aren't fast enough for that. Basically what makes processing a bunch of stand alone points more efficient than processing a bunch of polygons?

John Carmack has weighted in on it too apparently and see's it as something to potentially use in the 2-3 year future.

Notch (the minecraft guy) talks about it on his blog. And I think he's right in his assessment.
http://notch.tumblr.com/post/8386977075/its-a-scam
http://notch.tumblr.com/post/8423008802 ... not-a-scam

Rasterized graphics, ray casting and this voxel/point cloud technique all have their strengths and weaknesses. I personally think we'll see games using ray casting before we see anything mainstream and playable using this technology. But then it depends on if the right kind of hardware becomes mainstream.