Bluffmaster
Skilled
Is this the future of gaming?
Unlimited Detail Real-Time Rendering Technology
- Thread starter Bluffmaster
- Start date
Is this the future of gaming?
Bluffmaster
Skilled
Just think what it means for the future of Gaming? If what they are claiming is infact real then we could indeed see Avatar like graphics from the next generation of consoles.
saumilsingh
Skilled
Yea, these guys pop up every year with their fractal like procedurally generated 3D world.
Last year they showed some kinda alien jungle full of weird looking animals.
Notice how everything looks repetitive like the same objects are laid out side by side and on top of each other? That's how they manage to cram so many objects into the same scene.
Then they cleverly excuse it away by saying that they aren't artists.
Sure the sheer number of objects is high but when you sit down and count them, there's probably not more than 10-15 actual different objects making up the entire scene.
Sounds good in theory. Not so good for creating an actual vivid game world. Everything will end up looking like a weird Kaleidoscope of objects.
And unless they pull a rabbit out of their hats, I don't see it being any more impressive than it is right now, which is just cramming an obscene amount of the same objects into a single scene and pitching it as unlimited detail.
Last year they showed some kinda alien jungle full of weird looking animals.
Notice how everything looks repetitive like the same objects are laid out side by side and on top of each other? That's how they manage to cram so many objects into the same scene.
Then they cleverly excuse it away by saying that they aren't artists.
Sure the sheer number of objects is high but when you sit down and count them, there's probably not more than 10-15 actual different objects making up the entire scene.
Sounds good in theory. Not so good for creating an actual vivid game world. Everything will end up looking like a weird Kaleidoscope of objects.
And unless they pull a rabbit out of their hats, I don't see it being any more impressive than it is right now, which is just cramming an obscene amount of the same objects into a single scene and pitching it as unlimited detail.
Stuff like this has been used for more than a decade now, at varying performance levels obviously. Medical imaging and scientific simulations use such techniques all the time.
The differentiating factor, as saumil has pointed out is not just about cramping a bazillion objects into the scene (even when they are just repetitive instances). Video games need dynamic environments, not static scenery. Static scenery has existed for eons. Just cramming more instances of the same objects will not suddenly make the scene more dynamic.
Additionally, they talk about the whole 'atomish' point cloud based approach. Who is going to factor in storage for all the gazillion points involved? It has to be kept somewhere..
The differentiating factor, as saumil has pointed out is not just about cramping a bazillion objects into the scene (even when they are just repetitive instances). Video games need dynamic environments, not static scenery. Static scenery has existed for eons. Just cramming more instances of the same objects will not suddenly make the scene more dynamic.
Additionally, they talk about the whole 'atomish' point cloud based approach. Who is going to factor in storage for all the gazillion points involved? It has to be kept somewhere..
Gamespy interview with the Devs for those interested -
GameSpy: Unlimited Detail Interview - Page 1
--- Updated Post - Automerged ---
The article doesnt clarify but hints at how it handles memory so perhaps the same would apply to processing as well -
"Regarding the memory, If we were making our world out of little tiny atoms and had to store x, y, z, colour etc… for each atom, then yes it would certainly use up a lot of memory. But instead we've found another way of doing it. I could say we use less memory than what the current polygon system uses, but if I did that I think I'd exceeded my quote of unbelievable claims for the day. So well leave that for future demonstrations."
GameSpy: Unlimited Detail Interview - Page 1
--- Updated Post - Automerged ---
kvikram said:
The article doesnt clarify but hints at how it handles memory so perhaps the same would apply to processing as well -
"Regarding the memory, If we were making our world out of little tiny atoms and had to store x, y, z, colour etc… for each atom, then yes it would certainly use up a lot of memory. But instead we've found another way of doing it. I could say we use less memory than what the current polygon system uses, but if I did that I think I'd exceeded my quote of unbelievable claims for the day. So well leave that for future demonstrations."
A new, interesting interview. This tech may actually turn out to be very REAL rather than the hoax most devs are assuming it is. Time will tell....
HARDOCP - Euclideon & Unlimited Detail - Bruce Dell Interview - Euclideon & Unlimited Detail - Bruce Dell Interview
HARDOCP - Euclideon & Unlimited Detail - Bruce Dell Interview - Euclideon & Unlimited Detail - Bruce Dell Interview
challapradyumna
Skilled
What i Understood form the demo:- The current industry professionals are a bunch of retarded people who know to do things only the traditional way. We are super intelligent to get this kind of thinking.
My questions :- How do you texture a gazillion atoms and not take processing power?
Lets take it into 2D what he is saying is we took the vector graphics and made it into pixel graphics which takes way more less processing power and looks awesome.
Pulling a plain scene from a hard disk and rendering does not take much processing power. Rendering a scene with textures and that updates along with players actions that takes a s*** load of processing power. What they have shown is a static scene with no updating issues. What the industry needs is a dynamic scene with updating.
--- Updated Post - Automerged ---
Also the demo have repetitive objects which all shares the same texture once it is pulled from the harddisk the samething can be used for all the repeating objects but thats not the case with games.
My questions :- How do you texture a gazillion atoms and not take processing power?
Lets take it into 2D what he is saying is we took the vector graphics and made it into pixel graphics which takes way more less processing power and looks awesome.
Pulling a plain scene from a hard disk and rendering does not take much processing power. Rendering a scene with textures and that updates along with players actions that takes a s*** load of processing power. What they have shown is a static scene with no updating issues. What the industry needs is a dynamic scene with updating.
--- Updated Post - Automerged ---
Also the demo have repetitive objects which all shares the same texture once it is pulled from the harddisk the samething can be used for all the repeating objects but thats not the case with games.
saumilsingh
Skilled
Also that fact that our super ultra modern GPUs still can't render a simple sphere. To render a sphere, it has to put together millions of tiny triangles (polygons), which is a sheer waste of processing power.
So these atoms of theirs will still be composed of trillions of tiny triangles and you need some serious brute force to crunch them out onto the screen.
So these atoms of theirs will still be composed of trillions of tiny triangles and you need some serious brute force to crunch them out onto the screen.
saumilsingh said:
They don't use GPUs. Nor have the spoken about what primitives they use.
More importantly, they've still don't talk about what they are actually doing. IMHO the [H] interview doesn't really reveal anything new. It's still just the old airy fairy talk that's going on..