NVIDIA DLSS 5 Uses Generative AI to Enhance Game Realism; Huang Ren-hsun: A GPT Moment in Graphics Technology

NVIDIA CEO Jensen Huang announced the release of NVIDIA DLSS 5 at Monday’s GTC keynote, the latest version of the company’s AI graphics technology designed to make video game visuals more realistic while reducing computational resource consumption.

The new DLSS 5 system combines traditional 3D graphics data with generative AI models that can predict and fill in parts of images, allowing NVIDIA’s GPUs to generate detailed scenes and highly realistic characters without starting from scratch. NVIDIA claims DLSS 5 represents the most significant breakthrough in computer graphics since the introduction of real-time ray tracing in 2018.

DLSS 5 takes game color information and per-frame motion vectors as inputs and uses AI models to inject photo-realistic lighting and material effects into scenes. These effects remain anchored to the original 3D content and maintain consistency across frames. DLSS 5 operates in real-time at up to 4K resolution, ensuring smooth interactive gaming experiences. By bridging the gap between traditional rendering and the real world, DLSS 5 enables game developers to achieve unprecedented levels of realism previously only possible with Hollywood visual effects.

Huang said, “After 25 years since NVIDIA invented programmable shaders, we are redefining computer graphics once again. DLSS 5 is a ‘GPT moment’ in graphics. It combines human-designed rendering techniques with generative AI, achieving a huge leap in visual realism while maintaining artists’ creative control.”

“We are integrating controllable 3D graphics—the true benchmark of virtual worlds and structured data—with generative AI and probabilistic computing,” Huang stated. “The former is fully predictive, the latter probabilistic, but both can achieve extremely high realism.”

Huang noted that combining structured data with generative AI allows developers to create content that is both beautiful and stunning, yet controllable.

He added, “This concept of merging structured information with generative AI will repeatedly transform various industries. Structured data is the foundation of trustworthy AI.”

Although gaming now accounts for a smaller portion of NVIDIA’s overall revenue compared to the past, this industry laid the foundation for NVIDIA’s growth. Huang positioned DLSS 5 as part of a broader computational paradigm shift, implying its applications will extend far beyond gaming into enterprise-level computing.

Using enterprise data platforms like Snowflake, Databricks, and BigQuery as examples, Huang highlighted that structured data sets will become the basis for future AI systems to analyze and generate insights.

“The future is that these data structures will be widely utilized by AI, and AI will operate at speeds far beyond humans. Future AI agents will use both structured and unstructured databases, with the latter representing the vast majority of the world’s information.”

NVIDIA states that many top industry publishers and game developers, including Bethesda, CAPCOM, NetEase, Tencent, Ubisoft, and Warner Bros. Games, will integrate DLSS 5.

According to official announcements, DLSS 5 is expected to launch “this fall.” The first games supporting this technology include: “Eternal Tower 2,” “Assassin’s Creed: Mirage,” “Delta Force,” “Hogwarts Legacy,” “Reverse Water Cold,” “Naraka: Bladepoint,” “Otherworld,” “Starry Sky,” and “Yan Yun Shili Sheng,” among others.

(Source: Cailian Press)