I’m sure some who read this may have heard of a company called Euclideon who have produced a technology that allows for unlimited detail in computer graphics that can run on software alone with little gpu assistance. Their relatively secretive because they’ve received a ton of flak from many people about their technology, but i was hoping to bring this to light for the developers at I-Novae considering they’ve made a technology just as amazing. The videos on their YouTube page explain it much better than i can, but i’ll just say they are working on two games using their technology and are open licensing. I thought if you combined the ability of the I-Novae engine with the Unlimited Detail technology it open the game up to many more people and would definitely help its development.
I think it was similar topic i read on old forums, close before it went down. 
Don’t crash the site now please 
Where did you see that topic?
The old forums. They’re no longer accessible.
I remember few years back they boasted how they can render infinite worlds with super level of detail, but what ultimately came out of it is de-facto database system for data like 3d scans of whole areas. I know they demoed it few months ago, but result was crazy demanding demo (amusing considered low resource requirements was supposed to be selling point) that looked like NovaLogic returned from death with their voxel engine for one more kick:
Please get it right. It’s unlimited detail. Or is it UNLIMITED DETAIL? Maybe even UNLIMITED DETAIL. I think exclamation marks are optional. Annoyingly loud enunciation is not.
Euclideon has been engaged in point cloud data management for several years now - because that’s what they’re best at.
It is unfortunate that the public doesn’t have a lot of information on what they’ve been doing, other than their Geoverse technology they developed for the geospatial industry. They’ve had to go quiet many times because of all the scrutiny they’ve gotten, but even so i think their still worth looking into and considering in the long run. They’ve already shown in sources other than their demo that the technology is more than capable of being run on low end computers, depending on what you consider low end. And seeing how Infinity Battlescape already requires a hefty computer, helping reduce the performance rate while increasing consumer availability can only help it grow and develop. Isn’t that what we want? I’m just trying to help expose better possibilities.
i remember that years ago, i was wondering what happened with this, it’s interesting
I-Novae engine doesn’t work with voxels right now.
Technically, and i’m not saying you’re wrong, what they’re using is atoms which i’m guessing I-Novae doesn’t support either. How long would it take to implement that support however, because the point of Euclideons technology is to have an unlimited amount of them, one for each pixel on your screen.
That’s really just an (n+1)D voxel (where it’s+1 for the material), no?
Their technology is interesting, but I’m more interested in seeing something like open subdiv if we needed to increase our LOD to such extremes. https://www.youtube.com/watch?v=YATzaRDn_aQ
Their technology would matter only if they show dynamic real time lighting along with physics and animations and turn it into a game that runs in 60 fps else its all limited to point cloud voxel rendering. And when i say phyics/animations I mean that their tech should work with the following
1.) procedural animations.
2.) keyframe animations.
3.) physics based rigid body animations.
4.) soft body physics (cloth).
5.) fluid physics.
And Lighting has its own dynamics in rendering. So until and unless they achieve this in real time at over 60 fps with a running game as a demo their technology is basically limited and will fail, because they are claiming to bring a new rendering pipeline based on voxels instead polygons, but the whole world follows the traditional polygonal rendering approach because it works, whereas voxels have a ton of limitations at the moment due to performance etc.
Euclideon has some great tech and its interesting to see people discover it for the first time after being spread around the internet several times over the past 5 or so years.
I think this will be the FUTURE of graphics, atm they need to developer SDK kits in line to how good CryEngine and Unreal engine is. I do believe that is the main reason why we haven’t seen any real game or graphics system use it except for the geoscape system (pretty cool).
What will likely happen at some point once their confident with the UD engine is make their own game or benchmark. It’s just unlikely anyone with sane mind would invest development time with this engine until its been proven to have all the hallmarks of a fully functional next-gen graphics engine SDK. A graphics engine is only as good as its SDK I say.
If any developer must spend 20years making a decent game with a engine then it just isn’t worth it no matter how good it looks. You gotta be able to produce something within 3years these days, or have huge capital to last longer (almost no developers risk that with games these days).
I don’t think this will be anything. These point clouds are huge sets of data, cannot be easily modified, look horrible, don’t work with animations and don’t offer artistic freedom. There are already other conventional techniques available that look a hundred times better.
The technology is amazing for rendering point cloud data, but not for games IMO.
They actually use procedural technology to fill in the blanks in between the dots when user gets close to the data. Its not like conventional point cloud technology.
But you are right in saying animation MAY be a issue, or deformation, both of those have been shown off with the technology but in a very BASIC form which isn’t convincing as they need to show off high detail motion and such like a character talking and then going ragdoll, without the CPU dying (considering hundreds of these need to occur in games, sometimes many at once).
I won’t believe it until I see an interactive demo (like the ones that used to come with PhysX drivers) and an SDK
They’re working ( or already have it, haven’t checked for a while ) on animations. So I’m pretty sure they can/will do it.
IMO the real problem is that the entire world is static, and games have been moving away from an all-static world towards ones that are more dynamic. Physics, breaking stuff, and lots of little animations. How do you do trees/vegetation waving in the wind with this technology ? I’m pretty sure it’s close to impossible.
There are a lot of applications for this tech and it’s quite impressive, but I agree that games is probably not one of those.
If you can move a polygon with a picture or texture on it, why not a group of atoms? I won’t try and argue who’s right and who’s wrong, since it obvious it’s far from game engine standards and can’t be implemented into the I-Novae engine, but it’s still in development and worth keeping track of i think. I think Euclideon has done more than prove the technology is capable but just needs time. That’s why they’re pretty quiet; they don’t want to show things that aren’t ready and can be torn apart by others criticism.
You can’t move polygons within a mesh in a 3D renderer when it’s static either. Games use all sorts of tricks like kd-trees to optimize rendering, at which point the only way to animate a triangle is via shaders. Euclideon’s tech runs on the CPU so there are no hardware shaders available, and the quantity of atoms that you’d need to animate is much higher than triangles.