Home » Technology » NVIDIA’s Jetson Thor Powers Real-Time Reasoning in Robotics and Physical AI with Cutting-Edge AI Capabilities

NVIDIA’s Jetson Thor Powers Real-Time Reasoning in Robotics and Physical AI with Cutting-Edge AI Capabilities

by Sophie Lin - Technology Editor

“`html



<a data-mil="7854102" href="https://www.archyde.com/u-s-stocks-bleak-closing-day-three-major-indexes-surged-more-than-20-throughout-the-year-anue-juheng-us-stocks/" title="U.S. stocks' bleak closing day, three major indexes surged more than 20% throughout the year | Anue Juheng-US stocks">Nvidia</a> jetson Thor: A New Era of AI Robotics Dawns

SANTA CLARA, CA – August 26, 2024 – Robotics is on the cusp of a significant advancement with the widespread availability of Nvidia’s Jetson Thor modules. These cutting-edge robotics computers are set to become the central processing units for a new generation of robotic systems, impacting both research and commercial applications.

The Demand for Intelligent robotics

Modern robots require a substantial influx of sensor data and exceptionally low-latency AI processing capabilities. The operation of real-time applications demands significant computational resources to effectively manage concurrent data streams originating from numerous sensors.Jetson Thor addresses this challenge head-on, boasting 7.5 times more AI processing power, 3.1 times greater CPU performance, and double the memory of its predecessor, the Nvidia Jetson Orin.

Unlocking New Possibilities in Robotics

This substantial performance increase allows roboticists to process high-velocity sensor data and execute visual reasoning tasks directly on the device-capabilities that were previously too slow for practical request in dynamic real-world scenarios. This breakthrough unlocks exciting new possibilities for multimodal AI, notably in areas like humanoid robotics.

Industry Leaders Embrace Jetson Thor

Agility Robotics, a leading innovator in humanoid robotics, has already integrated Nvidia jetson technology into its fifth-generation Digit robot. The company plans to transition to Jetson Thor as the onboard compute platform for its sixth-generation model. this upgrade is expected to considerably enhance Digit’s real-time perception and decision-making abilities, enabling the robot to perform more complex AI-driven tasks. Digit is currently deployed in commercial settings, handling logistics operations such as stacking, loading, and palletizing in warehouses and factories.

“The enhanced edge processing provided by Jetson Thor will elevate Digit to the next level, improving its responsiveness and expanding its skill set to handle a wider range of complex tasks,” stated Peggy Johnson, CEO of Agility Robotics. “With Jetson Thor, we can deliver the latest physical AI advancements to optimize operations for our customers’ warehouse and factory environments.”

Boston Dynamics, renowned for its sophisticated robotic designs over the past 30 years, is incorporating Jetson Thor into its humanoid robot, Atlas. This integration will give Atlas access to server-level computing power, accelerated AI workloads, high-bandwidth data processing, and substantial memory-all within a compact, on-device package.

Beyond Humanoids: A Versatile Platform

The applications of Jetson Thor extend far beyond humanoid robots. It promises to accelerate growth in various fields, including surgical assistants, intelligent tractors, delivery robots, industrial manipulators, and visual AI agents, enabling real-time inference on the device for larger, more complex AI models.

A Leap forward in Real-Time Reasoning

Jetson Thor is specifically engineered for generative reasoning models. It empowers the next generation of physical AI agents, driven by large transformer models, vision language models, and vision language action models. These agents can operate in real-time at the edge, minimizing reliance on cloud connectivity.

Optimized with the Jetson software stack, Jetson Thor delivers the low latency and high performance required for real-world applications.It supports all prominent generative AI frameworks and AI reasoning models, including Cosmos Reason, DeepSeek, Llama, Gemini, Qwen, and domain-specific robotics models like Isaac GR00T N1.5, enabling seamless experimentation and local inference.

NVIDIA Jetson Thor real-time reasoning with multi-sensor input
NVIDIA Jetson Thor unlocks new capabilities for real-time reasoning with multi-sensor input. Further performance betterment is expected with FP4 and speculative decoding optimization.

Leveraging the Nvidia Cuda ecosystem throughout its lifecycle, Jetson Thor is anticipated to deliver increased throughput and faster response times with future software updates.

Comprehensive Software Support

Jetson Thor runs the complete Nvidia AI software stack, accelerating virtually every physical AI workflow with platforms including Nvidia Isaac for robotics, Nvidia Metropolis for video analytics AI agents, and Nvidia Holoscan for sensor processing.

These software tools streamline the development and deployment of applications, like visual AI agents that monitor worker safety via live camera feeds, humanoid robots capable of intricate manipulation tasks, and smart operating rooms that guide surgeons using data from multiple camera streams.

Driving Research Innovation

Research institutions, including Stanford University, Carnegie Mellon University, and the University of Zurich, are utilizing Jetson Thor to push the boundaries of perception, planning, and navigation models in a multitude of applications.

At Carnegie Mellon’s Robotics Institute, researchers are deploying Nvidia Jetson-powered robots to autonomously navigate complex, unstructured environments for medical triage, search, and rescue operations.

“The amount we can achieve is directly tied to the available compute power,” said Sebastian Scherer, Associate Research Professor and head of the AirLab at Carnegie Mellon University. “Previously, there was a significant gap between computer vision and robotics because computer vision tasks were too slow for real-time decision-making. Now, with faster models and computing, robots can handle more nuanced operations”.

Scherer expects that upgrading from existing NVIDIA Jetson AGX Orin systems to Jetson AGX Thor developer kits will enhance the performance of AI models like their award-winning MAC-VO model for robot perception at the edge, improve sensor fusion capabilities, and enable experimentation with robot fleets.

Jetson Thor Family and Availability

Component Description Pricing (USD)
Developer Kit Includes Jetson T5000 module, carrier board, heatsink, and power supply. $3,499
Jetson T5000 Module Production-ready module for integration into custom systems. $2,999 (1,000 units)

The Jetson ecosystem offers a comprehensive suite of tools and support for diverse application requirements, high-speed industrial automation protocols, and sensor interfaces, accelerating the time to market for developers. Partner companies like Advantech, Aetina, ConnectTech, Chiking, and TZTEK are producing production-ready Jetson Thor systems with adaptable I/O and custom configurations.

Sensor and actuator companies, including Analog Devices, e-con Systems, Infineon, Leopard Imaging, RealSense, and Sensing, are utilizing Nvidia Holoscan Sensor Bridge to seamlessly connect sensor data from cameras, radar, lidar, and other sources directly to GPU memory on Jetson Thor with ultra-low latency.

thousands of software companies can now enhance their existing vision AI and robotics applications with multi-AI agent workflows running on Jetson Thor. Leading adopters include Openzeka, Rebotnix, Solomon, and Vaidio.

The Nvidia Jetson AGX Thor developer kit is available for purchase now. The Nvidia DRIVE AGX Thor developer kit, designed for autonomous vehicles and mobility solutions, is available for preorder, with deliveries scheduled to begin in September.

frequently Asked Questions about Nvidia Jetson Thor

What is Nvidia jetson Thor?
Nvidia Jetson Thor is a powerful robotics computer designed to accelerate AI workloads at the edge, enabling more sophisticated and responsive robots.
What are the key benefits of Jetson thor?
Jetson Thor offers significant improvements in AI compute, CPU performance, and memory compared to previous generations, enabling real-time processing of complex sensor data.
What applications is Jetson Thor suitable for?
Jetson Thor is ideal for humanoid robotics, surgical assistance, smart agriculture, delivery robots, industrial automation, and visual AI applications.
What software platforms does Jetson Thor support?
Jetson Thor supports nvidia Isaac,Nvidia Metropolis,nvidia Holoscan,and a wide range of AI frameworks and models.
where can I purchase Jetson Thor?
Jetson Thor is available through Nvidia’s official store and authorized partners.

What impact do you foresee Jetson Thor having on the future of robotics innovation? And, how might these advancements shape the roles of humans in collaboration with robots?

Share your thoughts in the comments below and help us continue the conversation!

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Adblock Detected

Please support us by disabling your AdBlocker extension from your browsers for our website.