TheGreatWhiteShark
Member
Thanks
".....And to make up for the lack of information that comes from per-game networks, the company is making up for it by integrating real-time motion vector information from the game itself, a fundamental aspect of temporal anti-aliasing (TAA) and similar techniques. The net result is that DLSS 2.0 behaves a lot more like a temporal upscaling solution, which makes it dumber in some ways, but also smarter in others. "
"The single biggest change here is of course the new generic neural network. Looking to remove the expensive per-game training and the many (many) problems that non-deterministic games presented in training, NVIDIA has moved to a single generic network for all games "
at first glance it seems like instead of having one supercompute for each game separately, they created and are using a unified supercompute that they feed and use for all games.
so if I understood correctly, the supercompute is still a very crucial part for each and every game that offers DLLS. correct?
I probably read your comment wrong, and took it as if the supercompute part was not needed anymore.
Last edited: