Wouldn’t nvidia’s ai “ground truth” understanding of the game need to be retaught every time there’s an update to the game world? Compared to a static game that it only has to learn once?
1 Like
Minecraft has DLSS, so i don’t think they need to pre compute for every unique scene. I think it works with different combinations of textures and shapes or something. It’s very complicated.
It’s a good question but I don’t think it works like that because you could come up with the same argument for every single game that has DLSS right now. Games update all the time, release DLCs, all kinds of changes. From my understanding its now easier then ever for developers to use nvidia’s plugin now to add DLSS 2.0 to their games, its much more streamlined and easier to do then it was with DLSS 1.0 where developers had to litterately send coded samples direct to nvidia. I don’t believe this is the case now but who actually knows exactly? because they don’t disclose this, but its what I’m gathering from people like Linus who discussed it for like 10 seconds on one of their friday WanShow segments.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.