Recently, NVIDIA announced a new screen comparison video of DLSS 5 technology at the SIGGRAPH 2026 conference. In addition, Nvidia also announced more details. It has redone the problems that DLSS has previously criticized by players, such as false AI graphics, broken characters' facial features, and destruction of the game's native art.
Comparison video:
The videos were intuitively compared. Whether it is a close-up of a character's face or an indoor still life, DLSS 5 can retain the real texture without the blurring and deformation of the previous generation algorithm. The entire set of DLSS 5 generation technology does not require multiple cards. One RTX graphics card can complete the calculation, and ordinary gamers can fully experience the new features with their graphics cards.
NVIDIA will officially launch DLSS 5 this fall, and it will be open to major game manufacturers for adaptation. Developers can turn off AI enhancement for character faces individually and only enhance the effect for scenes and distant views. This not only preserves character details, but also relies on AI to improve environmental clarity and balance image quality and performance. The engine first outputs the complete original picture. DLSS 5 only optimizes materials, shadows, and reflections in the post-processing stage. The underlying geometry and lighting are completely unchanged, and there will be no AI arbitrarily changing the art style.
Video screenshot:
