NVIDIA has introduced the new DLSS 5 (Deep Learning Super Sampling) technology. The company promises to bring game visuals closer to photorealism by using the capabilities of the upcoming RTX 50-series graphics cards.
The technology uses an AI model to completely rethink lighting and materials in a scene.
It analyzes color data and motion vectors to “rebuild” realistic lighting while preserving the original geometry and textures of the game.
The neural network can now recognize objects by type. DLSS 5 processes skin, hair, water, and metal differently, applying specific effects to each. For example, it can use subsurface scattering for more realistic skin or complex reflections for metallic surfaces.
Frame generation has become a core part of DLSS 5. Since the lighting system is AI-based, almost every frame is partially generated by a neural network to achieve smoother performance, especially at 4K resolution.
However, DLSS 5 has already received criticism from gamers. The main complaint is that AI may ignore the work of artists and designers. Instead of showing the game exactly as intended, the technology can apply a unified “realistic” look that makes different games appear visually similar, like footage from a dashcam or ordinary photographs. Several well-known industry figures have also expressed concerns.
NVIDIA states that DLSS 5 is only an optional feature for players who want maximum immersion and realism, and that the technology will help deliver high-end visuals even on mid-range hardware.
Reminder: Team Heretics and Maysun sign multi-year partnership focused on health in esports
