NVIDIA’s GTC 2026 wrapped its final day in San Jose on Tuesday with a closing slate that extended well beyond Monday’s keynote announcements. The conference’s third day delivered Disney’s Kamino robotics platform in detail, a T-Mobile partnership for edge AI on cellular infrastructure, and continued updates to NVIDIA’s official GTC blog as late as 22:00 UTC.
Disney’s Kamino: Training Robots in Simulation at Theme Park Scale
The star of GTC 2026 — measured by audience reaction — was a three-foot robotic Olaf from Disney’s Frozen franchise. But behind the crowd-pleasing animatronic is a serious engineering platform.
Disney Research built Kamino, a GPU-accelerated physics solver that runs thousands of parallel reinforcement learning environments on a single GPU. The simulator trains robotic characters to navigate real-world conditions — uneven terrain, collisions with objects, and unstable surfaces — entirely in software before deploying them physically.
The practical test case: Olaf needed to perform in the Celebration in Arendelle show at World of Frozen in Disneyland Paris, which takes place on a boat in a lagoon. Through deep reinforcement learning in Kamino, the robot learned to balance on the moving vessel “in just a matter of hours,” according to Kyle Laughlin, SVP of Research & Development at Walt Disney Imagineering. That’s a task that would take a human child months.
Olaf’s hardware runs on NVIDIA Jetson for edge AI processing and NVIDIA Omniverse for simulation. Moritz Bächer, lab director of Disney Research, told a GTC breakout session that the goal goes beyond functional movement: “For us it’s not just about functional autonomy, it’s about believable autonomy. You want to believe that this is a character onstage and not just functionally going from A to B.”
Olaf debuts at Disneyland Paris on March 29. Disney signaled this is the beginning, not a one-off: “The speed at which we’re able to create new characters — and get them in front of our guests — is unprecedented,” Laughlin said. Disney Experiences posted $36 billion in revenue last fiscal year.
T-Mobile Brings Edge AI to Cell Towers
T-Mobile announced a partnership with NVIDIA to deploy AI processing directly on cellular infrastructure — cell towers and network equipment — rather than routing all AI workloads to centralized data centers.
The play targets a middle ground between cloud-based inference and on-device processing. AI tasks that require low latency — autonomous vehicle coordination, real-time traffic analysis, local agent execution — can run closer to users on NVIDIA hardware embedded in wireless infrastructure.
“T-Mobile is betting that its fast 5G network, accelerated by NVIDIA’s low-latency hardware, will open up more ways to handle the next wave of demanding data,” said CNET wireless carrier expert Jeff Carlson.
For the agent ecosystem, edge-deployed inference matters. It could reduce the round-trip latency that currently bottlenecks agent workflows relying on cloud APIs, particularly for time-sensitive tasks.
The Numbers Jensen Left On Stage
Huang reiterated NVIDIA’s forecast of at least $1 trillion in revenue from 2025 through 2027, driven by Blackwell and Vera Rubin system orders. The company described computing demand as having “increased by 1 million times over the last few years,” fueled by $150 billion in venture capital flowing into AI-native startups.
GTC 2026 hosted 450+ sponsors, 1,000 sessions, and 2,000 speakers across its three days. Partner announcements from ABB, Universal Robots, KUKA, Hyundai, BYD, Nissan, and Geely continued filing through Tuesday.
What GTC 2026 Actually Delivered
Across three days, NVIDIA’s conference produced a coherent five-layer thesis:
- Silicon: Vera Rubin platform (Vera CPU, Rubin GPU, BlueField-4 STX), with the next-generation Feynman architecture and Rosa CPU announced as the follow-on.
- Models: Nemotron Coalition rallying six frontier model families with partners across language, vision, robotics, driving, biology, and climate.
- Agent infrastructure: NemoClaw and OpenShell for enterprise-grade OpenClaw deployment, with Jensen Huang declaring that “every single company in the world today has to have an OpenClaw strategy.”
- Physical AI: Disney Kamino, robotaxi partnerships with BYD and Uber, and industrial integrations with ABB, Universal Robots, and KUKA.
- Edge and scale-out: T-Mobile edge AI, DSX Air for simulating AI factories before building them, and the NVIDIA Space-1 announcement for orbital AI data centers.
The Monday keynote set the vision. Days 2 and 3 filled in the execution detail. With Disney deploying Kamino-trained characters at Disneyland Paris in ten days and NemoClaw positioning NVIDIA as the enterprise backbone for OpenClaw, GTC 2026 was the conference where NVIDIA declared the agentic AI era operational — and backed it with shipping dates.
Sources: NVIDIA GTC Blog, CNET Live Blog, BizTech Magazine, WDW News Today, Disney Experiences, TechRepublic