What Happened
During a press conference, a conversation with NVIDIA employees provided clarification about the upcoming DLSS5 technology. According to the account, the employees explicitly stated that DLSS5 makes no changes to character models at all, does not alter game assets, and does not apply any form of "AI filter" to characters.
The employees attributed the perceived visual differences in demonstrations or previews solely to changes in lighting. They emphasized that the underlying 3D models and textures remain the same, unaltered assets. This clarification appears to address community speculation or concerns that DLSS5 might be using generative AI techniques to modify character appearance.
Context
DLSS (Deep Learning Super Sampling) is NVIDIA's AI-powered upscaling and frame generation technology. Each major version has introduced new capabilities:
- DLSS 2: Focused on image reconstruction and upscaling.
- DLSS 3: Introduced Frame Generation for GeForce RTX 40-series GPUs.
- DLSS 3.5: Added Ray Reconstruction for improved path-traced lighting.
Speculation about DLSS5 has centered on what new AI-driven graphical enhancements it might bring. This employee clarification directly counters one line of speculation—that it might alter character models—and reinforces that its improvements are likely in the domain of lighting and image reconstruction, consistent with the technology's trajectory.
Source: Account from @kimmonismus on X/Twitter detailing a conversation with NVIDIA employees at a press conference.
