Become a CyberRobo Pro member to unlock more
Subscribe now
01. NVIDIA Officially Launches Jetson Thor, Accelerating the Local Implementation of Humanoid Robots, Leading to Potential Explosive Growth in Edge Devices and Computing Modules
NVIDIA has finally launched its long-awaited Jetson Thor GPU chip designed for embedded AI applications. This chip allows for local operation, reducing reliance on cloud services and significantly lowering the latency from perception to decision-making. NVIDIA is now distributing Jetson Thor to partner companies such as Agility, Fourier, Figure, and Diligent Robotics.
Specifically, the next-generation NVIDIA Jetson Thor platform is designed for physical AI and humanoid robots, supporting any popular AI framework and generative AI models, and is fully compatible with NVIDIA’s software stack from cloud to edge, including NVIDIA Isaac for robot simulation and development, the Isaac GR00T humanoid robot base model, NVIDIA Metropolis for visual AI, and NVIDIA Holoscan for real-time sensor processing. This platform expands NVIDIA’s ecosystem of over 2 million developers, providing edge AI deployment support for more than 7,000 customers across various industries.
With this humanoid robot-specific chip, several key technologies and development costs that hinder the large-scale implementation of humanoid robots can be significantly reduced.
Therefore, the Jetson Thor chip truly brings large models, complex perception, and real-time decision-making into the humanoid robot itself, eliminating the need for cloud processing, resulting in low latency and efficient execution in practical operations (enhanced multi-sensor, low-latency fusion capabilities, etc.). This means humanoid robots become more practical, further promoting their large-scale implementation.
As humanoid robots enter mass production and deployment applications, the demand for Jetson Thor chips is expected to explode in the future.
Jetson Thor-level edge computing capabilities shift tasks previously only possible in the cloud to the device itself, directly creating demand for higher bandwidth I/O, more memory, LPDDR/HBM modules, power supply/PMIC, cooling modules, motherboards/SoMs, and high-performance cameras and connectors. Orders for these components in the supply chain will subsequently increase.
Additionally, NVIDIA’s platform effect (Isaac/Omniverse + a large number of developers/partners) lowers the integration threshold, prompting more OEMs, system integrators, and sensor manufacturers to quickly follow suit, thereby expanding procurement of standardized edge computing modules and reference designs. The more active the ecosystem, the greater the downstream hardware demand.
On the market level, multiple industry reports and market forecasts indicate that the edge AI chip/device market is in a rapid growth phase (CAGR in double digits), providing a macro growth backdrop for component manufacturers, ODM/EMS, and sensor suppliers.
It is worth mentioning that during the World Robot Conference, Professor Wang He from Galaxy General announced that Galaxy General’s G1 is the first in China to apply NVIDIA’s Jetson Thor chip, a GPU chip specifically designed for humanoid robots. If you saw their demonstration on-site, you would notice that the G1 has significantly lower latency and faster operation compared to previous technologies, helping it complete tasks such as retail grabbing, logistics handling, and sorting…
The full text is 3,330 words, providing periodic intelligence/information interpretation, hoping to inspire subscribers beyond just information.
Become a CyberRoboPro member to read moreSubscribe now: exclusive insights, Pro community discussions, business cooperation, etc.
Next, Pro members can read, welcome to join CyberRoboPro
-
In-depth report on the commercialization strategy and market ecology of intelligent robots | CyberRoboPro
-
Customer votes with their feet: Pragmatism and ultimate vision in the humanoid robot industry | RoboInsight July
-
Humanoid Robots’ Explosive Month: Soft and Hard Integration, World Models, and the Game of Wheeled Bipedalism | RoboInsight 52
-
Humanoid Robots Have Many Wildernesses, Finding That North Star | RoboInsight June
-
Avoiding the Narrative of the Universe Tree | RoboInsight 51
-
The Collective Effect of Small Humanoid Robots, Opportunities for Humanoid Robots in the Construction Industry | RoboInsight 50
-
Popping the Bubble, Delving into Scenarios: Who is Truly Winning the Future of Humanoid Robots? | RoboInsight 49
-
Wheeled Precedence and the Revival of Dalu | RoboInsight 48
-
Humanoid Robots Have Many Wildernesses, Finding That North Star | RoboInsight June
-
Optimus Going to Mars, Scenarios Reign Supreme, Humanoid Robot “Nuclear Fusion” Quietly Occurring | RoboInsight May
-
The Frenzied Financing Narrative, Agility’s Milestones, and UBTECH’s Liuzhou Model | RoboInsight 47
-
Subscribe to the premium version of ChatGPT, and receive a humanoid robot? When NVIDIA uses humanoid robots to create AI, from 4S stores to JD.com | RoboInsight 46
-
Sell Yourself or Sell Your Skills? The Humanoid Robot 4S Store is Here, Figure is Selling Work; Some Manufacturers are Selling Products | RoboInsight 45
-
Entering the Experience Era: Humanoid Robots Begin to “Accumulate Life”; Two Fronts: “Quick Money” and “Long-termism” | RoboInsight 44
-
G1’s Fighting Competition, a Heavy Blow to the Brains of Domestic Humanoid Robots | RoboInsight 43
-
Commercialization of Humanoid Robots is Diversifying: From Factories to Hotels, Elderly Care; Capital Continues to Pour In, Over 1.9 Billion Yuan Flowing to Humanoid Robot Manufacturers | RoboInsight March
-
Humanoid Robots in Commercial Use: Interaction and Effectiveness Determine Success, Beware of the Loss of Value in the Body Under the “Component Frenzy” | RoboInsight 42
-
Modern Group’s Dual-Line Layout, the 1000N2’s松延动力 Opens Up the “Xiaomi Model”, Is Youliqi’s Chinese Version 1X? | RoboInsight 41
-
The Service Robots that Exploded at the Canton Fair, the IP of Humanoid Robots, and the Acceleration of VLA for Family Applications | RoboInsight 40
-
Automakers Leading the Humanoid Robot Industry and Implementation, Small but Deep Vertical Opportunities and Batch Applications of Service Humanoid Robots | Robo Insight 39
-
Forget the Half Marathon, Focus on the Collaborative Dexterous Hands and Humanoid Robots in Hotels, Elderly Care, and Agricultural Use Cases | RoboInsight 38
-
Japan and South Korea Catching Up to China and the US, the Cold and Hot Applications of Humanoid Robots | RoboInsight 37
-
Elon Musk’s Return, New Developments in Optimus; Agility’s New Product Strengthens Commercialization; Q1 Financing Exceeds 11.2 Billion | RoboInsight 36
-
Daniel Kahneman’s System 1 and System 2 are Still Rising in Value, Household Humanoid Robots are Getting Closer | RoboInsight 35
-
Humanoid Robot Component “Free-for-All”, Dumbbell-Shaped Market Emerges; Figure’s Ambition | RoboInsight 34
-
Collective Collaboration and Expansion, Figure’s Valuation Soars, Ranking Surpasses OpenAI | RoboInsight 33
-
Collective Intelligence of Humanoid Robots Begins, 7 New Products and the Formation of a Global Humanoid Robot Network | RoboInsight February
-
A Victory for Bionic Tendon Humanoid Robots, Giants Continue to Enter the Field, Regional Competition for Humanoid Robots Kicks Off | RoboInsight 31
-
The Richest Companies in the World are All Investing in Humanoid Robots | Robo Insight 30
-
100 Companies Influencing Global Humanoid Robots, Humanoid Robot Dining Shops and Apple’s “Non-Verbal” Humanoid Robots | RoboInsight 29
-
The AI Humanoid Robot Arms Race Between China and the US; European Dark Horses and Manufacturing Giants’ Ambitions and the Need for a Baymax | RoboInsight 28
-
The Struggle of Kings, Batch Delivery and New Growth Curve Dilemmas, and the Unignorable HRI Applications | Robo Insight 27
-
New Wealth in 2025: 11 Humanoid Robots Emerge, Over 5.8 Billion in Financing and New Opportunities in Vertical Markets | RoboInsight January