Key Announcements at GTC

Nvidia’s GPU Technology Conference opened with CEO Jensen Huang delivering a nearly three‑hour keynote that underscored the company’s role as the backbone of modern artificial intelligence. The event highlighted three major themes: physical AI through robotics, the rise of agentic AI platforms, and ambitious plans to extend data‑center capabilities beyond Earth.

Disney Robotics Partnership

One of the most eye‑catching moments was the reveal of a collaboration with Disney that brought the beloved snowman from the film "Frozen" to life as a robot. The Olaf droid runs simulations on Nvidia GPUs and is powered by Nvidia chips, using the Newton Physics Engine—a joint effort by Nvidia, Google DeepMind, and Disney Research—to enable high‑performance robot simulations. The partnership suggests that similar robotic installations could appear in Disney theme parks.

Agentic AI Platform – NemoClaw

Nvidia introduced NemoClaw, a stack built on the OpenClaw agent platform that lets developers create autonomous AI agents, or “claws.” The toolkit integrates the Nvidia AI Agent Toolkit, OpenShell for open‑model support, and a sandbox environment that adds layers of privacy and security. Huang described the OpenClaw event as a milestone comparable to HTML or Linux, positioning NemoClaw as a foundational tool for the next generation of AI‑driven applications.

Space‑Based Data Centers – Vera Rubin

The keynote also featured a teaser for Vera Rubin Space‑1, a concept for data centers operating in orbit. Huang explained that cooling in space relies solely on radiation, presenting unique engineering challenges. The Vera Rubin platform aims to help AI companies build and deploy tools in environments where traditional convection and conduction are unavailable.

Focus on Inference Workloads

Huang emphasized that the current inflection point in AI is inference—the process by which trained models apply knowledge to new data. He noted that inference now drives the majority of compute demand, especially for agentic AI that must continuously adapt. Nvidia claims to offer the lowest cost per token, positioning the company as a “token king” in the emerging inference economy.

Additional Highlights

The event also previewed DLSS 5, an AI‑powered upscaling technology that adds element‑based understanding to improve realism in games. While specific GPU support was not confirmed, the feature is expected to align with Nvidia’s upcoming Blackwell architecture. Discussions of new consumer‑focused chips, such as the rumored N1 and N1X, hinted at a possible return to the Windows laptop market through a partnership with MediaTek.

Overall, Nvidia’s GTC keynote painted a picture of a company extending its AI leadership across hardware, software, and even outer‑space infrastructure, while forging high‑profile collaborations that bring AI into everyday experiences.

Used: News Factory APP - news discovery and automation - ChatGPT for Business