Nvidia DLSS 5 Slammed by Gamers and Developers
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- DLSS 5, Nvidia's upcoming "real-time neural rendering model" slated for Autumn 2026, uses generative AI and a game's internal color and motion vector data to overhaul lighting and textures rather than just upscale resolution; CEO Jensen Huang said it merges "generative AI" with "handcrafted rendering"
- The DLSS 5 demo currently requires two RTX 5090 GPUs, with one card entirely dedicated to running the neural rendering model
- Digital Foundry called the "transformational lighting" effects "astonishing" in its detailed breakdown, but the broader gaming world reacted overwhelmingly negatively, with Nvidia's YouTube trailer comments filling with thousands of negative takes
- Game developers including Mike Bithell (_Thomas Was Alone_), Gunfire Games Senior Concept Artist Jeff Talbot, and New Blood Interactive CEO Dave Oshry publicly attacked the technology, saying it strips away artistic intent and amounts to a "garbage AI filter"
- Nvidia responded in those YouTube comments that DLSS 5 "is not a filter" and that "game developers have full, detailed artistic control," including the ability to mask the effect off entirely in certain scenes
Why it matters: Nvidia's pitch of AI-driven re-lighting crossed a line beyond the upscaling gamers have tolerated since 2018's RTX 2080, and the industry's near-unified pushback—including from Gunfire Games' senior concept artist and New Blood Interactive's CEO—suggests the Autumn 2026 rollout will be developer-by-developer, not blanket. The demo's two-RTX-5090 requirement signals a halo feature, not a mainstream upgrade.
