Galbot, a global leader in embodied intelligence and robotics, has integrated NVIDIA Jetson AGX Thor into its G1 Premium robot, recently unveiled at the World Robotics Conference (WRC). As one of the industry’s first adopters of Jetson Thor, Galbot is boosting performance by enabling faster, smoother, and more intelligent autonomy, setting a new standard for deployments across retail, healthcare, and logistics.
With the integration of NVIDIA Jetson Thor, Galbot has significantly raised the bar for general-purpose robotics, accelerating its roadmap and demonstrating how edge computing enhances autonomy and commercial deployment. The Jetson Thor-powered G1 Premium delivers significant advances in speed, flexibility, and real-time inference capabilities. With 7.5x the AI compute performance and 3.5x the energy efficiency of its predecessor, NVIDIA Jetson Orin, Galbot’s robots redefine the possibilities of embodied intelligence by executing precise planning and motion.
AI Authority Trend: Beamr Shows 50 Percent Better Video Compression for Autonomous Vehicles with NVIDIA
“Powered by NVIDIA Jetson Thor, G1 Premium delivers significant improvements in speed and real-time inference performance, enabling our proprietary VLA models to perform even better in real-world environments,” said Professor Wang He , founder and CTO of Galbot . G1 Premium demonstrated outstanding performance at WRC, earning recognition as the “swiftest humanoid worker.”
Building on this foundation, Galbot, in collaboration with Tsinghua University and Shanghai Qi Zhi Institute, jointly developed OpenWBT_Isaac. This is a specialized simulation platform for whole-body teleoperation of humanoid robots. It utilizes the whole-body control function of the OpenWBT system and virtual-reality fusion technology based on NVIDIA L20•RTX 5880 Ada GPUs to enable the development of efficient and advanced humanoid robots.
AI Authority Trend: Broadcom and NVIDIA Expand Partnership for AI on VMware Cloud
At the heart of Galbot’s success lies its proprietary Sim2Real methodology. This method pretrains on a large scale using a large, high-quality synthetic dataset, then refines it with minimal real-world data. This drastically reduces reliance on costly real-world data collection while significantly improving generalization performance across complex environments.
Galbot also demonstrated its technological prowess at the World Humanoid Robot Games, winning first place in every round from preliminaries to the finals, operating fully autonomously without remote control. Powered by the proprietary Sim2Real methodology and a large-scale VLA model, Galbot’s G1 robot is already operating fully autonomously in over 10 pharmacies in Beijing, with plans to expand to over 100 locations nationwide by the end of the year. Furthermore, Galbot is collaborating with leading manufacturers, including the Bosch Group, to deploy its AI-based platform in smart production, replacing traditional robotic arms with more intelligent and adaptable solutions.
AI Authority Trend: Nvidia, Microsoft, And the Future of AI Stocks Amid Market Volatility
Source – PR Newswire
To share your insights, please write to us at sudipto@intentamplify.com
