Laut Li Mu Bai/ B3D, der behauptet, einen guten Draht zu einem prominenten Wii U Entwickler zu haben, sollen meine Theorien zutreffend, und die GPU der Wii U ein Hybride aus Unified Shader und Fixed Function Architektur sein. Insbesondere das Lighting soll Nintendo dabei priorisiert haben:
http://forum.beyond3d.com/showpost.p...&postcount=833Zitat von Li Mu BaiFirst off, this board is far too educated to presume that developing a console GPU is like shopping at Newegg. Or even determining current component prices, if it was AMD/ATI, Genyo Takeda (IR&D) would not have spent 2yrs.+ in the development process on the Wii U's graphics processor. I realize we are attempting to establish a power as well as an architectural baseline, but this will be an amalgamation of processor capabilities that will yield a very custom proprietary chip. Somewhat defying the current DX metric. (to a degree of course)
What I mean when I say that is this, we cannot assume because it's based off of, or similar to gpu architecture "X," that it is incapable of "Y." Y equaling effects such as tessellation, IBL, real-time GI, deferred rendering, etc. There are certain visual aspects, such as lighting, that are very important to Nintendo. I have heard that, much like the Flipper, Nintendo has incorporated at least partially a portion of the same design philosophies into the Wii U chipset. Features that “automagically” appear during shader code implementation. A post from my early days regarding the GC’s architecture on B3D:
"However, as mentioned above, a couple of features where added in automagically already, like self-shadowing and tinting for example."
"Per-object self-shadowing can be realized quite nicely on the Nintendo Gamecube. The benefit of doing self-shadowing on a per object basis is that one does not need to be concerned so much with precision."
"One should note that during the shader build many features are activated dynamically. For instance, if an object should get tinted a color multiplication is added to the final output color whatever shader was setup before."
"The results of global lighting can be computed in three different ways: per vertex, per pixel using emboss mapping, and per pixel using bump mapping. All three of these methods come in two variants one with self-shadowing and one without."--Florian Sauer & Sigmund Vik http://www.gamasutra.com/features/20.../sauer_pfv.htm
Also 8 light values came at a very negligible performance cost, because Flipper computed light values in parallel to UV generation. It’s these types of “hardwired” like effects Nintendo I believe has carried over to make modern shader effects with a subset of fixed feature functionality . I’m simply providing examples as I do not know to what extent overall it is, or can be incorporated. (esp. with the gpu being of a modern design) I was told that lighting behaved in this manner, & that lighting was a point of emphasis. As always with a secondhand source, you must always be cautious not to take it as gospel. (though I trust this source, Nintendo's NDAs are the most binding)
Nintendo did make certain alterations to their gpu based upon various 3rd party input, a first. Usually, they tend to develop their gpus & platforms with just simply ATI/Nintendo engineering, consultation, & guidance. Designed around their evolving software strengths, & "the natural flow of the industry."-Genyo Takeda Yes, I am referring to all those benchmark tests Nintendo ran on 3rd party engines for optimization on Wii U hardware.
But make no mistake, Nintendo's footprint is definitely here. You will see a marked performance difference in their proprietary engines, as well as close 3rd parties, & exclusive titles. (UbiSoft, Capcom, etc.) Also, ARM may also be providing their DSP component solution. The nameless devs that are claiming inferiority to the current generation of consoles are either inept, or working with middleware that is still yet unoptimized for the differing Wii U architecture.