NVIDIA unveiled DLSS 5 today, describing it as the most significant breakthrough in computer graphics since the debut of real-time ray tracing in 2018. The technology introduces a real-time neural rendering model capable of delivering photoreal lighting and materials inside games, achieving a level of visual quality that until now was only possible in Hollywood visual effects production. Jensen Huang, founder and CEO of NVIDIA, called it “the GPT moment for graphics,” describing it as a combination of hand-crafted rendering with generative AI that delivers a dramatic leap in visual realism while preserving the control artists need for creative expression.
DLSS 5 works by taking a game’s color and motion vectors for each frame as input, then using an AI model to infuse the scene with photoreal lighting and materials anchored to the game’s 3D content and consistent from frame to frame. The system runs in real time at up to 4K resolution for smooth, interactive gameplay, and the AI model is trained to understand complex scene elements such as characters, hair, fabric and translucent skin, along with environmental lighting conditions like front-lit, back-lit or overcast scenes, all from analyzing a single frame. From there, it generates visually precise images that handle details like subsurface scattering on skin, the sheen of fabric and light interactions on hair, while retaining the structure and semantics of the original scene.
To give developers full creative control, DLSS 5 includes detailed settings for intensity, color grading and masking, so artists can determine exactly where and how enhancements are applied without compromising each game’s unique visual identity. Integration is handled through the same NVIDIA Streamline framework already used by existing DLSS and NVIDIA Reflex technologies, making adoption straightforward for studios already familiar with the pipeline.
A decades-long pursuit of photorealism in real time
The challenge NVIDIA is trying to solve with DLSS 5 has been at the core of real-time graphics for decades. A Hollywood VFX frame can take minutes or even hours to render, while a game running at 60 frames per second only gets about 16 milliseconds per frame, a gap that raw computing power alone cannot close.
NVIDIA has been chasing that gap since the programmable shader arrived with GeForce 3 in 2001, then CUDA with GeForce 8800 GTX in 2006, real-time ray tracing with GeForce RTX 2080 Ti in 2018, and path tracing with neural shaders on the GeForce RTX 5090 in 2025, accumulating a total increase of 375,000x in compute power over that span. Even with all of that progress, the rendering budget of a 16-millisecond game frame remains a fraction of what’s available to a photoreal movie frame.
DLSS has been part of that effort since its release in 2018, starting as an AI technology to boost performance through resolution upscaling and eventually generating entire frames. It has since been integrated into over 750 games and become an industry gold standard. DLSS 4.5, launched at CES earlier this year, already used AI to draw 23 out of every 24 pixels visible on screen. With DLSS 5, the focus shifts beyond performance into visual fidelity, using AI not just to fill in pixels faster, but to make them look fundamentally better.
Major publishers and titles already on board for this fall
NVIDIA announced that DLSS 5 will be supported by major publishers and developers including Bethesda, CAPCOM, Ubisoft, Warner Bros. Games, Tencent, Hotta Studio, NetEase, NCSOFT and S-GAME. The confirmed game lineup includes Resident Evil Requiem, Assassin’s Creed Shadows, Starfield, Hogwarts Legacy, The Elder Scrolls IV: Oblivion Remastered, Phantom Blade Zero, Delta Force, NARAKA: BLADEPOINT, AION 2, and more titles to be announced.
Todd Howard, studio head and executive producer at Bethesda Game Studios, said that seeing DLSS 5 running in Starfield amazed the team and that they can’t wait for players to experience it. Jun Takeuchi, executive producer at CAPCOM, said the technology represents an important step toward the cinematic, deeply believable visuals the studio strives for in Resident Evil.
Charlie Guillemot, co-CEO of Vantage Studios, said that the way DLSS 5 renders lighting, materials and characters changes what they can promise players in Assassin’s Creed Shadows.
DLSS 5 was previewed at GTC during the week of March 16, with demos running in Resident Evil Requiem, EA SPORTS FC, Starfield, Hogwarts Legacy and the NVIDIA Zorah tech demo. The technology is set to arrive in games this Fall. The before-and-after comparisons already shown publicly make the difference hard to ignore, skin, fabric and hair all respond to light in ways that blur the line between a game and a filmed production.
What do you think about DLSS 5, is AI-powered graphics the future you’ve been waiting for? Tell us in the comments!