Joe DeFuria said:
In the DX8 era? (Same API as X-Box 1). Absolutely.
Saying the same "API" is wrong, very wrong due to intrinsic differences between programming methodologies that are used on the open PC verse the closed nature of a Console. I don't know of many developers that programmed to DX8 on XBox - most are a lot lower than that.
Something you didn't mention though, which there is no doubt in my mind, is that the concept of programming to "Shaders" was expedited due to Microsoft's almost cannibalization of PC gaming. In effect, XBox creating an artificial plateau which PC developers were confident that programming to would be economically viable. This then trickled back down to the PC ports which targeted the DX8 level, and as you somewhat stated was advantageous to nVidia.
The difference between
Then and
Now (something you didn't cover) is that DX8 was still catering to a large amount of fixed functionality and rigid hardware. Thus, having someone work exclusively with your microprogram/shader models was a huge boost as there was this big differential between each IHV's implementations of these more of less fixed functions.
DX9+ will see the creation of a nearly complete instruction set. The days of having this "feature" or that "feature" are quickly dying - and with it are many (not all) of the same perks of having 'your' implementation in a console. There will be no large differentials between supporting N hardware or M hardware that invoke the use of API dependent extensions or specific IHV centric DX derivatives that can be leveraged "back" to the PC arena - you'll support this one instruction set and it'll be standard. By the time the NGConsoles launch in 2005, I expect the industry to have moved in this direction in a big way*.
* As in, way past the NV3x’s DX8/DX9 duality that caused problems. We’ll be much further down the rabbit hole of offering computation power en masse and tapping it via extreme extraction – which is pretty hardware indifferent when all is said and done.
Bouncing Zabaglione Bros. said:
"OK, one thing I have to say about ATI and Nintendo Cube. Did the Cube deal help ATI corner the high-end market? Then why was R3x0 so good?
Also, you're overlooking the fact that unlike the XGPU, the Flipper chip didn't share any commonality with PC parts. It's a custom chip on a DX7 level that has 'perks'. nVidia got the XBox boost because programmers could become accustomed to their shading models and then port a game back to the PC with a high degree of ease when targetting the more advanced Nv25 level and PC that existed (eg. more RAM, CPU, IO, etc).