DLSS 5: Nvidia Presents 3D-Guided Neural Rendering
An in-depth look at the controversial AI technology
The strong negative reaction that accompanied the announcement of DLSS 5 last March must have caught Nvidia by surprise. The US multinational intended to convey the message that it had created a revolutionary technology in the field of video game rendering, but instead, its tool was perceived by many observers as an inappropriate intrusion of artificial intelligence into game design itself.
We entrusted our thoughts on Nvidia's communication to a dedicated special report. Now the company returns to talk about DLSS 5, delving into neural rendering with a new presentation video. Edward Liu and Gabriele Leone of Nvidia explain in detail the challenges this new technology had to overcome, how it works, and the possibilities it offers to developers.
However, the controversy has left its mark, which is clearly evident from how the discussion is framed in the video: it is repeated over and over again that DLSS 5 does not alter the original geometries, so as not to affect the author's artistic vision. To reinforce the concept, examples with tech demos are brought out. It is also emphasized that developers will have accurate tools available to fine-tune DLSS 5, so they can achieve a result in line with their vision.