Personally I do not think it is a good idea to expect any of these things universally. AI? This is game Design more than computational complexity - Fear and halo 1 are ancient yet most fps still uses call of duty AI types that have very low dynamism and little to no systemic behaviours. That is unless you just mean the shere amount of npcs on screen or in memory. Regarding the rest of all the features, no game need to use them all at once - as project Design makes that a complete waste. Not every game Design needs constant streaming, mesh shaders, or etc. To me what makes a next gen game is doing just one aspect important to the game's existence that is not feasible in a real time manner or at playable quality on a previous Generation of technology.
But I expect big AAA openworld to use some of the functionnality at least Streaming, replace mesh/primitive shader little by little, having vastly improved dynamic lighting. I can't imagine how GTA 6 will look like. This is a matter of budget, some game cost 100 to 200 millions dollars to make.
One 343 industry dev think you can replace geometry rendering pipeline little by little games after games.
http://reedbeta.com/blog/mesh-shader-possibilities/
The other really neat thing about mesh shaders is that they don’t require you to drastically rework how your game engine handles geometry to take advantage of them. It looks like it should be pretty easy to convert most common geometry types to mesh shaders, making it an approachable upgrade path for developers.
(You don’t have to convert everything to mesh shaders straight away, though; it’s possible to switch between the old geometry pipeline and the new mesh-shader-based one at different points in the frame.)
The funny thing Nvidia are basically two years having all new features. Out of RT where the Nvidia advantage is so big, people could have made the R&D on Turing GPU but they were not knowing what AMD can deliver.
Last edited: