Predicting the Xbox-2

WELL now that i think about it, high polygons counts are useful not only to represent things u actually see on screen...

i mean, all those polygons or 3d calculation will be used to display loads of realistic looking and acting shadows, or things that affect the environment even though they are not visible on screen....

also, i you want a realistic looking tree, or a park full or trees, with individually rendered leaves, will require quite a lot of polygons....

hair, realistic looking hair, will require quite a lot of polygons, and with all this talk about billions of polygons per second, the 60.000 individual hairs on Aki's head (in FF:TSW) seem like nothing...

u know what i mean... there are LOADS of things u'll be able to do with those polygons... displacement maps increases poly counts as well...
 
Some stuff on Microsoft research.



Geometric Models


Research in this field focuses on the surface representation of complicated geometric objects for the purpose of quick and accurate rendering.

Multiresolution geometry

Highly detailed geometric models are becoming commonplace in computer graphics. These models, often represented as complex triangle meshes, challenge rendering performance, transmission bandwidth, and storage capacities. We have introduced progressive meshes, which capture complex models as coarse meshes together with refinement streams. We develop accurate simplification metrics for forming such progressive meshes. The goal is to preserve not just the geometry of the original mesh but, more importantly, its overall appearance as defined by appearance attributes. We also research real-time rendering of large-scale surfaces through view-dependent refinement.


Subdivision Surfaces
We have researched methods for automatic reconstruction of accurate piecewise-smooth subdivision surface models from scattered range data. The method can be used in a variety of applications such as reverse engineering the automatic generation of CAD models from physical objects. We have also studied displaced subdivision surfaces, which offer a number of benefits, including geometry compression, editing, animation, scalability and adaptive rendering.


Surface Textures
We have produced a method for creating texture over an arbitrary surface mesh using an example two-dimensional (2-D) texture by identifying interesting regions (texture patches) in the 2-D example and repeatedly pasting them onto the surface until it is completely covered. We call such a collection of overlapping patches a lapped texture.


Efficient Rendering
Rendering detailed surface models requires many triangles, resulting in a geometry processing bottleneck. We are working on efficient rendering by clipping the rendering of coarse geometry to the exact silhouette of the original model. In addition, we are investigating the use of vertex caching to transparently reduce geometry bandwidth across the graphics bus.

Non-Photorealistic Based Rendering

Escherization

Here is the Escherization problem: Given a closed figure in the plane, find a new closed figure that is similar to the original and tiles the plane. Our solution works by using an optimization technique to search over the 93 different types of "isohedral" tilings. Isohedral tilings are a class of tilings in which each tile is equivalent to every other. These tilings, it turns out, are flexible enough to encompass nearly all of Escher’s own tilings, and yet simple enough to be encoded and explored by a computer. This is joint work with the University of Washington.

http://research.microsoft.com/graphics/



In the near future, Microsoft's virtual sports games may soon have more realistic crowd scenes thanks to Salesin and fellow researchers' development of "video textures," a type of videotape loop that creates mini-movies that don't appear to repeat. He is also helping develop ways to create more realistic, three-dimensional human faces on computers and to train computers to work as virtual cinematographers for PC movies and video games.

http://www.microsoft.com/presspass/features/2000/jul00/07-26salesin.asp

Tiling Layered Depth Images
This image represents work by the Graphics group to model scenes that capture the complexity of the real world. The group has developed what is known as Layered Depth Images (LDIs). By extending the notion of an image to contain multiple colors and pixels at each point a viewer can move his or her head back and forth in the scene.Previously, landscapes have been successfully rendered as static images, but resemble Astroturf in interactive worlds--the grass will not wave as you “walkâ€￾ past, the sunflowers look painted on. The Graphics group has now developed a technique to render complex three dimensional scenes by using a sophisticated tiling method for LDIs. Each sunflower tile holds on average 20 sunflowers and each sunflower is comprised of 35,000 triangles.

http://research.microsoft.com/features/images.asp
 
A lot of neat stuff here about Xbox2. Anyways, I want to add that I think the XB2 will use a dual CPU solution on a single chip. I've heard that no one's is really creating a totally new chip design to use the newer processes in the future but many are thinking about putting multiple CPUs on a single chip.
 
MS has the ability to go alone(or maybe work with Bitsboy) and carry on with their Talisman project. Is it not true that PS2 rendering philosophy kinda follow up on the Talisman project?
 
Why stop at dual ?

Doesn't those R350 or NV35 capable of more ?

I think MS will see the spec for PS3 and chained together as many GPU or CPU together to beat it. They only care about having monopoly in the industry, and with their resources it is within reach.
 
Yep, sure MS can have it to meet(and surpass?) what the competitors are doing. Didnt someone said that 3DO took on Nintendo 64 Reality Hype and delivered the killer M2(that surpasses the actual N64)? I am sure MS will take that 1TFlops or whatever info and aimed to surpass that.
 
chaphack said:
Yep, sure MS can have it to meet(and surpass?) what the competitors are doing. Didnt someone said that 3DO took on Nintendo 64 Reality Hype and delivered the killer M2(that surpasses the actual N64)? I am sure MS will take that 1TFlops or whatever info and aimed to surpass that.

aren't you worried that innovation is being buried under the weight of all those dollars?
 
Depends. MS will have it out to battle. But either way, i think you cant get much innovation without the weighty dollars! ;)
 
chaphack said:
Depends. MS will have it out to battle. But either way, i think you cant get much innovation without the weighty dollars! ;)

while I agree that without investment innovation becomes stagnent, don't you think that any entity with that much money and influence will even bother to do so?
 
Sure they will.

MS(same with Sony and Nintendo) will try their best to outspend and outinnovate the rest. Until the time comes when one guy(MS, Sony or Nintendo) beats the hell, and i mean BEAT THE HELL!!! outta the other guys(MS, Sony or Nintendo), then maybe innovation might slowdown. ;)
 
chaphack said:
MS has the ability to go alone(or maybe work with Bitsboy) and carry on with their Talisman project. Is it not true that PS2 rendering philosophy kinda follow up on the Talisman project?

Your joking right?
 
I am not joking, i am asking. ;)
Wasnt fast framebuffer/brute software rendering gotta something to do with either Bitsboy or Talisman :?: :?: :?:
 
Talisman was almost an image based / compositing system chaphack.
The idea was to reduce the workload by 'warping' bitmaps rather than recalculating everything every frame ( like quicktimeVR or matrix bullet time stuff )
PS2 is almost more like SGI infinity engine - with fully programmable geometry processor ( especially with the ability to generate more vertex out than vertex in data ) but without the complex per pixel logic found in the PC graphics cards.
 
notAFanB said:
aren't you worried that innovation is being buried under the weight of all those dollars?

No, If it wasn't for the XBox, Sony might not be aiming so high with PS3. Right now there is HEAVY competition on both the PC and console platform, which is driving technology forward incredibly fast. It's great!
 
Josiah said:
notAFanB said:
aren't you worried that innovation is being buried under the weight of all those dollars?

No, If it wasn't for the XBox, Sony might not be aiming so high with PS3. Right now there is HEAVY competition on both the PC and console platform, which is driving technology forward incredibly fast. It's great!


good point.
 
Crazyace said:
Talisman was almost an image based / compositing system chaphack.
The idea was to reduce the workload by 'warping' bitmaps rather than recalculating everything every frame ( like quicktimeVR or matrix bullet time stuff )
PS2 is almost more like SGI infinity engine - with fully programmable geometry processor ( especially with the ability to generate more vertex out than vertex in data ) but without the complex per pixel logic found in the PC graphics cards.

In that case, i guess i need more reading up on Talisman. :p
 
notAFanB said:
Josiah said:
notAFanB said:
aren't you worried that innovation is being buried under the weight of all those dollars?

No, If it wasn't for the XBox, Sony might not be aiming so high with PS3. Right now there is HEAVY competition on both the PC and console platform, which is driving technology forward incredibly fast. It's great!


good point.

Hey! That was what i said too. :p
 
Back
Top