Xenos Texture Filtering Quality

[Luminescent]

According to Dave's article on Xenos:

Each of the filtered texture units have Bilinear sampling capabilities per clock and for Trilinear and other higher order (Anisotropic) filtering techniques each individual unit will loop through multiple cycles of sampling until the requested sampling and filtering level is complete. The texture address processor has some general purpose shader ability and is able to apply offsets from the input texture co-ordinates which can be used with higher order filtering techniques. The Anisotropic filtering capabilities adapts the number of samples taken dependant on the gradient of the surface that it is sampling, which is fairly normal for Anisotropic filtering mechanisms, ATI says that the anisotropic filtering quality is improved from previous generations of hardware. As Xenos is the controller of a UMA, the entirety of system RAM is available to the texture samplers, although they will not perform any operations on the eDRAM memory.

So the Aniso alorithm should be of higher quality than in ATI's current solutions and the texture address units should allow for filtering techniques of a higher order. What would it take, then, for Xenos to display shimmer free texture filtering with crisp texture detail (considering the effects of distance on the perception of detail) while frame perspectives are in motion? Would it absolutely require higher levels of aniso to alleviate this (not a usual solution for console apps) or just a solution that samples completely on an adaptive, case by case basis, perhaps based on what is specified by the app?

I noticed that in Crytek's Far Cry demo for ATI, there was very little shimmering, even on Nvidia's current GPUs, when settings were set to application preference and settings filtering set to high quality in the control panel. Were they shaders that made the difference in filtering quality or was it just the fact that viewing was done from a fixed perspective?

Aside from AnisoWhat higher order techniques would be viable performance wise and how would they improve upon current anisotropic solutions?

 

[Megadrive1988]

I dont get it. why couldn't Xenos have been made so that each texture unit produces a trilinear filtered sample each clock. doesn't the ArtX Flipper do trilinear filtering each clock without loop ?

 

[OpenGL guy]

I dont get it. why couldn't Xenos have been made so that each texture unit produces a trilinear filtered sample each clock. doesn't the ArtX Flipper do trilinear filtering each clock without loop ?

I don't know anything about Xenos, but if you are primarily combining textures, then you want fast trilinear. However, if you're primarily executing shaders, then texture filter speed becomes less important.