This is DLSS 5 and its commitment to neural imaging

If you play on PC, it’s very likely that you’ve already come across three letters that are appearing more and more often in the graphics settings of games: DLSS. In recent years, this NVIDIA technology has become one of the parts that best explains why GeForce cards can offer higher performance or higher resolution without requiring the same effort from the hardware. Now the company has decided to go even further. As part of GTC 2026, it unveiled DLSS 5, a new evolution of its technology that it says aims not only to improve performance or image scaling, but also to change the way the final image is generated on screen using what it describes as a real-time neural rendering model.

Advertising. DLSS 5 goes a step further than what we’ve seen so far in this family of technologies. The company describes this new version as a neural rendering model that adds photorealistic lighting and materials to the scene. To achieve this, the system analyzes data generated by the game itself in each frame, such as color and motion vectors, and applies a trained artificial intelligence model to understand the semantics of the scene. This training allows the system to interpret complex elements, from the behavior of light on hair or fabric to effects such as light scattering on skin, while preserving the original structure of the game’s 3D world.

How Nvidia Dlss 5 Scaled 2 works

Change focus in DLSS. Until now, the conversation around this technology has almost always focused on performance. It usually had to be activated to increase the fps or to be able to play at higher resolutions without demanding as much from the GPU. With the announcement of DLSS 5, NVIDIA presents a different nuance. The company explains that its new neural rendering model is not limited to reconstructing the image created by the game, but can change its visual appearance, including more complex lighting and materials, with the intention of bringing real-time graphics closer to the level of fidelity traditionally associated with visual effects in cinema.

Nvidia Dlss 5 On Off 1

The role of AI. The DLSS 5 proposal is based on an artificial intelligence model trained to interpret the semantics of the scene appearing on the screen. NVIDIA explains that the system takes as input the color and motion vectors of each frame and from there creates lighting and materials that remain stable between frames and are tied to the game’s original 3D data. This training allows the model to recognize complex elements of the scene, from the interaction of light with hair or fabric to effects such as light scattering on skin. At the same time, the company emphasizes that developers can adjust parameters such as intensity, color gradation or the areas where these improvements are applied, and claims that all this can be done in real time at up to 4K.

Discussion and criticism. The presentation of DLSS 5 was not without controversy. After the publication of the official video, some experts in the sector and users of social networks criticized that the technology significantly modifies the original appearance of some games. Some of the criticism centers on AI adding lighting and photorealistic materials that can alter artistic decisions made during development. Faced with these challenges, NVIDIA emphasized that studios will be able to adjust parameters such as intensity, color or areas where the effect is applied to maintain the visual identity of each work.

The video game industry is doing the unexpected: Xbox is following in Sony's footsteps and backing away from the idea of ​​a PC console

Big question. If you’ve been following NVIDIA news closely, it’s likely that this announcement has you wondering. As a reminder, DLSS 4.5 was presented at this year’s CES. Now, shortly after that, comes DLSS 5. What we’re seeing here isn’t necessarily an immediate replacement, but rather a new commitment from NVIDIA to push its graphics technology into a different realm. As for the calendar, NVIDIA has planned the initial deployment of DLSS 5 for the fall of 2026. The company assures that the technology will receive the support of several major studios and publishers in the sector.

Images | NVIDIA

In Xataka | If the question is how Nintendo makes money, the answer isn’t in video games: it’s a much more ambitious company

Leave a Reply

Your email address will not be published. Required fields are marked *