Nvidia reveals artificial intelligence system capable of multiplying graphics performance by up to six times
The technology conference held recently at Las Vegas served as the stage for one of the most anticipated announcements in the hardware and digital entertainment industry. Nvidia has officially announced the arrival of version 4.5 of Deep Learning Super Sampling (DLSS), an image resizing tool that promises to establish new performance paradigms for home computers. The major innovation presented lies in the integration of a second generation transformer model, specifically developed to maximize visual fluidity in games that demand high computational processing power.
The central point of the technical presentation was the new system’s unprecedented ability to multiply frame generation, providing much greater stability for end users. The company confirmed that the super resolution functionality is now available through the official application for all cards in the RTX line. However, the full dynamic generation feature, which employs sophisticated algorithms to create intermediate frames, is scheduled to launch in the spring, initially being exclusive to RTX 50 series equipment.
Currently, the manufacturer’s ecosystem covers a vast library, exceeding the mark of 400 titles compatible with super resolution technology. Desse in total, around 250 games already offer native support for frame generation, a number that should grow exponentially as the integration of the new tool is simplified for development studios. The goal is to popularize the use of artificial intelligence in graphic rendering, relieving the load on the graphics processor and delivering visual results that often surpass traditional native rendering.
The technological leap of transformative models
The DLSS 4.5 architecture is based on a completely renewed transformer model, which was trained with a significantly larger volume of data and has greater processing capacity compared to previous versions. Essa structural evolution allows image scaling with high precision, correcting visual flaws that were common in past generations of technology. The main focus of development was eliminating distortions in fast-moving objects and improving the definition of small light sources, ensuring clarity even in the most frantic action scenes.
More on this story: Nvidia may launch RTX 50 Super series in 2026 and include 12GB RTX 5060 Super
Company experts highlight that performance mode is now capable of achieving visual quality equivalent to or even greater than native 4K resolution. Esse advancement is crucial for players who use maximum graphics settings, including full ray tracing. Neural optimization means enthusiasts no longer have to choose between game fluidity and visual fidelity, uniting performance and image quality through AI-assisted processing.
The technical improvements implemented in this new version attack specific points that directly influence the user’s immersion during gaming sessions:
– Redução drastic visual artifacts, known as “ghosting”, in fast-moving objects.
– Aumento noticeable in the sharpness of luminous details, volatile particles and complex atmospheric effects.
– Otimização dedicated to monitors with extremely high refresh rates, exceeding the 240 Hz barrier.
– Significant Melhoria in the readability of user interfaces (UI) and small texts at high resolutions.
See Also
Learn more: New generation of NVIDIA RTX 50 SUPER graphics cards expected to debut only at CES 2027
Frame multiplication and visual fluidity
The big competitive advantage of version 4.5 is the frame generation feature with a multiplier of up to six times. The system operates intelligently, creating up to five additional frames through artificial intelligence for each frame traditionally rendered by the GPU. A proprietary algorithm adjusts this multiplier in real time, constantly analyzing hardware workload and scene complexity to ensure system stability.
During moments of intense action and in visually charged scenarios, the multiplier works to keep the frame rate high. On the other hand, in less demanding situations, the intervention is modulated to preserve latency and the immediate response of commands. Essa adaptive approach is essential for titles that use path tracing, where the demand for raw processing is intense, enabling smooth gameplay even at 4K resolutions.
Preliminary tests indicate that the technology can maintain the responsiveness of the controls even with the high frame rate generated artificially. Isso solves one of the most recurring criticisms of the first iterations of frame generators, which often introduced noticeable delays in commands (input lag) in exchange for a more fluid image, which hindered the experience in competitive games.
Availability and market impact
The practical application of these innovations can now be observed by users who activate the feature through the manufacturer’s application settings. The update fixes clarity issues in interface elements, something that affected readability in complex games like “The Outer Worlds 2”. Nvidia promises that integration by developers will be simplified, eliminating the need for extensive and complex updates to the codes of games that are already on the market.
Hardware owners of the upcoming RTX 50 series will have exclusive initial access to 6X dynamic generation mode, marking an important generational leap in the industry. Contudo, the commitment to the installed base remains firm, ensuring that the second-generation transformer model works on all generations of supported RTX cards. Isso ensures that even users with previous generation hardware will notice immediate gains in sharpness and visual noise reduction in their favorite games.
On the same topic: Video card prices rise drastically with Asus and Gigabyte announcing new adjustments in 2026
The continuous expansion strategy reinforces the position of artificial intelligence as the central pillar of modern graphics development. With more than 650 applications and games already using some form of the technology, version 4.5 arrives to consolidate AI-assisted rendering as the industry standard, extending the useful life of components and enabling next-generation visual experiences for an increasingly wider audience.
