TLDR: Nvidia has launched its Jetson AGX Thor platform, integrating the advanced Blackwell GPU architecture into robotics for the first time. This move significantly boosts AI computing power and energy efficiency for industrial and humanoid robots, marking a new era of ‘Physical AI’ where machines can learn and interact with the real world.
Nvidia is ushering in a transformative era for robotics, dubbed ‘Physical AI,’ with the introduction of its Jetson AGX Thor platform. This latest offering marks a pivotal moment as it integrates Nvidia’s cutting-edge Blackwell GPU compute architecture directly into both industrial and consumer robotics, including advanced humanoid systems. Traditionally, Nvidia’s robotics platforms have lagged behind its most advanced GPUs, with the previous Jetson Orin platform based on the older Ampere architecture. However, with the Jetson AGX Thor, Nvidia has aligned its robotics offerings with its most powerful AI compute engine, ensuring that every one of its vertical segments now leverages its latest chip technology.
The performance enhancements are substantial due to this generational leap. Nvidia reports that the Jetson T5000 production module delivers an impressive 7.5-times increase in AI computing power and a 3.5-times improvement in energy efficiency compared to its predecessor, the Jetson Orin. These significant gains are expected to unlock a new range of applications, particularly enabling the development of humanoid-style robots suitable for commercial deployment and, eventually, consumer markets.
Central to the Jetson AGX Thor platform is the new Blackwell Multi-Instance GPU (MIG) technology. This feature allows the GPU to be partitioned into multiple virtual segments, enabling the system to rapidly and consistently handle a diverse array of robotic activities and sensor inputs. Beyond hardware, Nvidia is also expanding its robust robotics software ecosystem. The Nvidia Jetson software platform has been updated to fully support these new boards, and the company is integrating powerful tools such as its Isaac Groot humanoid robot foundation models and Nvidia Metropolis for Vision AI.
Nvidia’s vision for ‘Physical AI’ involves applying the algorithmic principles used in large language models (LLMs) for text-based interactions to enable robots to learn and replicate physical movements in the real world. Rev Lebaredian, Vice President of Omniverse and Simulation Technologies at Nvidia, emphasized this convergence, stating, “Computer graphics and AI are converging to fundamentally transform robotics.” Nvidia CEO Jensen Huang has frequently described Physical AI as the ‘next wave’ of artificial intelligence, following the eras of Perception AI, Generative AI, and Agentic AI, with the key distinction being the ability of robots to sense, reason, and interact physically with their environment.
Also Read:
- Nvidia’s Automotive Sector Accelerates with 69% Revenue Growth Driven by Self-Driving Innovations
- NVIDIA’s Vision: The Dual Factory Future Driven by Physical AI and Autonomous Systems
Early adoption of Nvidia’s Jetson Thor technology is already evident, with major robotics companies such as Figure, Amazon, and Caterpillar utilizing the platform. While robotics currently represents a smaller portion of Nvidia’s overall revenue, accounting for approximately 1%, the segment has demonstrated significant growth, with a reported 72% quarterly sales increase. This growth trajectory, coupled with Nvidia’s strategic investments and technological advancements, positions the company to be a dominant force in the rapidly expanding robotics market, where machines are increasingly expected to not only think but also act in the physical world.
