[Theory] Next frontier of graphics: solid models?

Currently everything rendered in games are basically shell wireframes, so only the visible layer of the object is rendered and it has nothing inside (no skeleton or anything). This requires complex formulas to simulate realistic movement an animation. That's why we have such a hard time making people look realistic. I think Nvidia is barking up the wrong tree with adding imperfections such as pockmarks, because who wants to see Lara Croft with blackheads?

My proposal is to simulate the entire body. Skeleton, muscles, skin, lungs, blood, the works. All movement would be defined with how these elements interact, and damage (such as cuts or bruises) can be rendered accurately. And attacking would be more realistic because instead of requiring multiple shots to the head to kill an enemy, you'd have to hit an actual vital spot.

This, of course, requires much more powerful gaming systems, and could be what AMD and Nvidia need to keep the threat of Intel's "good enough" away.


Thoughts?

Cool idea but will prolly never happen. Not only would you need a monster PC but the games would be horrendously expensive due to the sheer amount of programmers required.

It might be enough to fire up the cpu doubling law again.