Nvidia CEO Jensen Huang used the company’s GTC keynote to outline a major shift in artificial intelligence computing, introducing DLSS 5, a new AI-powered graphics system, while projecting unprecedented demand for the company’s next-generation chips. The updated DLSS technology combines structured 3D graphics data with generative AI models to predict and render visual elements, enabling more realistic scenes while reducing compute requirements.
Huang indicated that this fusion of deterministic data and probabilistic AI represents a broader evolution in computing, with potential applications extending beyond gaming into enterprise systems. He pointed to platforms such as Snowflake, Databricks, and BigQuery as examples of structured data environments that future AI systems could leverage alongside generative models.
In parallel, Huang highlighted the accelerating demand for Nvidia’s AI hardware, projecting that combined orders for its Blackwell and Vera Rubin chip architectures could reach $1 trillion through 2027, up from approximately $500 billion previously anticipated. The Rubin architecture, designed to outperform Blackwell, is expected to deliver significantly higher performance in both model training and inference.
The announcements underscore Nvidia’s central role in the expanding AI economy, spanning both software innovation and the infrastructure powering next-generation AI systems.
