| zarx said: I will just let Carmack explain
|
I think John Carmack is saying something similar to what I was saying but our expectations of what this means may be different ...
Graphics processing power can be used to create more detailed graphical assets (models, textures, etc.), to increase the quality of the lighting and effects, or to improve the quality of the output image (increased resolution, AA, AF). To a certain extent you can improve all of these in isolation, but eventually you will hit a point where you need to improve the others to see any real benefit; as an example, if you were using an output resolution of 320x240 it would be difficult to see the difference between a game like Resident Evil 4 on the Wii/Gamecube and Resident Evil 5 on the HD consoles because the capabilities of these systems both exceed what can (really) be displayed at such a low resolution.
Now, if you had 10+ times the processing power of the HD consoles you can focus on "maxing out" the detail that can be displayed at 720p @30fps and the quality of lighting/effects that can be produced using the raster scan line algorithm, or it could be used to render at 1080p @ 60fps with more minor improvements to detal. The lower resolution approach would probably result in a system that seemed to be more of a jump forward in graphics, but the higher resolution approach would likely keep development costs lower.
If you could get to a level that was 40+ times the processing power of the processing power of the HD consoles you could probably increase resolution and detail at the same time but, being that would probably require graphical hardware that is the equilivalent of 2 of the top of the line graphics cards that will be released in 2012, and that means the system would either be released much later or be very expensive.
