NVIDIA today introduced DLSS 5, marking its most significant advancement in computer graphics since real-time ray tracing’s launch in 2018. DLSS 5 employs a real-time neural rendering model that imbues pixels with photorealistic lighting and materials, closing the gap between rendered graphics and reality. This enables game developers to achieve Hollywood-level visual effects in interactive environments. Jensen Huang, NVIDIA’s founder and CEO, described DLSS 5 as a “GPT moment for graphics, ” blending traditional rendering with generative AI to dramatically elevate visual realism while preserving artistic control. Since the GeForce series began, NVIDIA has consistently advanced graphics technology—from programmable shaders in 2001, to CUDA in 2006, real-time ray tracing in 2018, and now path tracing with neural shaders in 2025—delivering a 375, 000x boost in computational power. Despite this, real-time gaming frames have far less rendering horsepower than time-intensive Hollywood VFX frames, meaning brute force alone cannot achieve photorealism in games. First launched in 2018, DLSS initially enhanced performance by upscaling resolution and later generating new frames, becoming an industry standard integrated into over 750 games. DLSS 4. 5, unveiled at CES earlier this year, uses AI to render 23 of every 24 pixels in real time. DLSS 5 advances this further by enriching visual fidelity alongside performance. Unlike offline video AI models—which generate photoreal images but lack real-time control and consistency—DLSS 5 inputs each frame’s color and motion vectors into an AI model that renders photoreal lighting and materials anchored to the original 3D content, maintaining frame-to-frame consistency.

It supports real-time gameplay at up to 4K resolution. The AI model is trained end-to-end to comprehend complex scene elements—such as characters, hair, fabric, translucent skin—and various lighting conditions by analyzing single frames. DLSS 5 then generates precise visuals, capturing subtle effects like subsurface scattering on skin, fabric sheen, and light-material interactions on hair, while preserving the scene’s structure and semantics. Developers receive detailed controls for intensity, color grading, and masking to tailor the enhancements and maintain each game’s unique look. Integration is streamlined through NVIDIA’s existing Streamline framework used by DLSS and Reflex technologies. Major publishers and studios—Bethesda, CAPCOM, Hotta Studio, NetEase, NCSOFT, S-GAME, Tencent, Ubisoft, and Warner Bros. Games—support DLSS 5. Bethesda’s Todd Howard praised its impact on Starfield, calling it “amazing” and transformative. CAPCOM’s Jun Takeuchi highlighted DLSS 5’s role in crafting cinematic, believable experiences in Resident Evil, enhancing atmosphere and emotional depth. Charlie Guillemot, co-CEO of Vantage Studios, noted DLSS 5’s contribution toward immersion in Assassin’s Creed Shadows by elevating the realism of lighting, materials, and characters, enabling the creation of more ambitious worlds. In sum, NVIDIA DLSS 5 redefines real-time graphics by leveraging advanced AI models to achieve photorealism once exclusive to Hollywood, enabling unparalleled visual fidelity and immersive gaming experiences.