Nvidia’s DLSS 5 introduces impressive AI-driven enhancements in lighting and facial detail that could revolutionize gaming visuals but currently demands high-end hardware and exhibits issues like artifacting and unnatural effects. While promising significant performance and quality gains, its proprietary nature risks fragmenting the GPU market, potentially sidelining competitors and limiting accessibility unless Nvidia opts for greater openness.
Nvidia recently unveiled DLSS 5 at GTC 2026, a neural rendering technology that has sparked a wave of memes and mixed reactions online. While some scenes showcased during the demo appeared jarring and gave rise to criticism, the technology itself promises significant improvements in global lighting and facial detail in games. The presenter acknowledges the impressive visual enhancements but criticizes the trend of locking such advancements behind expensive proprietary hardware, forcing gamers to upgrade frequently to access the latest features. Despite the backlash, DLSS 5 could dramatically shift the hardware landscape, potentially sidelining AMD and Intel in the discrete GPU market.
The technology’s impact is evident in games like Assassin’s Creed Shadows and Starfield, where lighting and character faces show marked improvements. However, the presenter points out that some of these gains could have been achieved through traditional graphics optimization if developers had invested more effort over the past decade. The facial improvements in Starfield and FIFA demonstrate DLSS 5’s potential when used subtly, but there are also instances, such as a character in Hogwarts Legacy, where the AI-generated enhancements look unnatural and unsettling. Performance-wise, DLSS 5 currently demands two RTX 5090 GPUs for real-time demos, raising concerns about its accessibility until the upcoming RTX 6000 series arrives.
Artifacting and ghosting remain issues with DLSS 5, as seen in demos like Resident Evil Requiem and Starfield, where distracting visual glitches undermine the technology’s benefits. The presenter criticizes the AI’s tendency to over-enhance scenes, leading to overly sharp and contrast-heavy visuals that can feel artificial. Interestingly, DLSS 5’s neural rendering could render Nvidia’s existing ray tracing and path tracing features obsolete, as it offers superior visuals and potentially better performance. This shift conveniently aligns with Nvidia’s interest in promoting new GPU sales, as current hardware may struggle to run DLSS 5 effectively.
Looking ahead, the presenter speculates that DLSS 5 will require developer support and likely won’t see widespread adoption until later this year with the RTX 6000 series launch. There are concerns that the technology might encourage developers to be lazier in optimizing games for conventional hardware, relying instead on AI to enhance visuals. While this could free resources for gameplay innovation and storytelling, it might also lead to job cuts in traditional graphics roles. The next generation of Nvidia GPUs, possibly based on the Reuben CPX GPU, is expected to excel in AI workloads, making DLSS 5 and similar features central to future gaming experiences.
A major point of contention is the potential exclusivity of DLSS 5 to Nvidia hardware. Unlike previous features that AMD and Intel could replicate, the AI models behind DLSS 5 are proprietary and trained by Nvidia, meaning competitors would have to develop their own versions, resulting in inconsistent visual experiences across hardware. This fragmentation could harm the gaming community by creating divergent looks for the same game depending on GPU brand. The presenter calls for Nvidia to open-source DLSS 5’s models to ensure a unified gaming experience and warns that failure to do so might mark the end of meaningful competition in the GPU market, with negative consequences for gamers and PC enthusiasts alike.