iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
At a London “Build a claw” event reported on October 15, 2026, NVIDIA CEO Jensen Huang presented OpenClaw as a model for agentic computing that could extend from software into physical machines. He compared a robot’s motors, wheels, articulation and arms to tools controlled by an agent running inside a computer, while warning that autonomous systems need strict isolation, minimum privileges and continuous oversight before broad deployment.
What Jensen Huang said about OpenClaw and robotics
TechRadar’s Mike Moore reported that Huang called OpenClaw “one of the most important models” and connected agentic systems with coding, marketing assistance, research, laboratory work, drug discovery and engineering experiments. Huang was speaking at a London “Build a claw” event on October 15, 2026.
The robotics analogy came after Huang viewed a robot built by NVIDIA partner Humanoid. He described OpenClaw as the “first credible” agentic harness and said a robot can be understood as the physical harness of a large language model: its motors, articulation, wheels and arms are tools used by an agent sitting inside the computer.
That framing does not mean OpenClaw has independently demonstrated safe control of robots. It describes a software architecture: an agent decides what to do, while a physical platform supplies the actuators and sensors needed to carry out tasks. The cited event report provides no independent benchmark, incident-rate statistic or regulator finding establishing that OpenClaw is safe for autonomous robotics.
#1 Best Overall
- The NVIDIA Jetson AGX Orin 64GB Developer Kit makes it easy to get started with Jetson Orin. Compact size, lots of connectors, and up to 275 TOPS of AI performance make this developer kit perfect for prototyping advanced AI-powered robots and other autonomous machines.
- The developer kit includes a Jetson AGX Orin 64GB module, and can emulate all the Jetson Orin modules. It supports multiple concurrent AI application pipelines with the NVIDIA Ampere GPU architecture, next-generation deep learning and vision accelerators, high-speed IO and fast memory bandwidth. Now you can develop solutions using your largest and most complex AI models to solve problems such as natural language understanding, 3D perception, and multi-sensor fusion.
- Jetson runs the NVIDIA AI software stack, and use-case specific application frameworks are available, including Isaac for robotics, DeepStream for vision AI, and Riva for conversational AI. You can save significant time with NVIDIA Omniverse Replicator for synthetic data generation (SDG), and by using NVIDIA TAO toolkit to fine-tune pretrained AI models from the NGC catalog.
- Jetson ecosystem partners offer additional AI and system software, developer tools, and custom software development. They can also help with cameras and other sensors, as well as carrier boards and design services for your product.
- With the computing capability of more than 8 Jetson AGX Xavier systems in a developer kit that integrates the latest NVIDIA GPU technology with the world’s most advanced deep learning software stack, you’ll have the flexibility to create tomorrow’s AI solution as well as today’s.
Why Huang says controls must come first
Huang argued that autonomous agents should not receive unrestricted access to a computer’s underlying hardware or operating system. He compared the need for containment with early internet browsers that could reach bare metal, then described OpenShell as “essentially another operating system inside the operating system.”
His operating rule is zero trust: assume the agent could make an unsafe choice, give it only the rights required for the assigned task, observe its behavior continuously and provide a way to escalate or stop actions that conflict with policy.
Huang illustrated the principle with a kitchen example. If an agent is asked to make an omelet, it needs access to a frying pan, not a chainsaw. In technical terms, an agent that can prepare a meal should not automatically be able to access unrelated tools, devices, files or system functions.
- Isolation: Run the agent in a contained environment separated from bare metal and the host system.
- Zero-trust assumptions: Treat every requested action as untrusted until policy permits it.
- Minimum privileges: Grant only the tools, data and permissions needed for the current task.
- Continuous monitoring: Check actions against policy while the agent is operating, rather than relying only on a one-time approval.
- Escalation controls: Route uncertain, dangerous or non-compliant actions to a human or a higher-level security system.
Huang summarized the deployment pattern as “zero trust, minimum policy, continuous monitoring, contained inside an isolation system that is separated from bare metal.” His stated goal is to make autonomous machines and agents useful without giving up the ability to constrain or interrupt them.
Rank #2
- Complete Jetson Orin Nano Starter Kit: This jetson orin nano starter kit includes a 30-in-1 sensor board, 8MP camera, dual-servo gimbal, 128GB SD card, and essential accessories. It supports Avisual recognition and voice interaction, providing a complete AI application development experience
- 8MP AI Vision Camera with Gimbal: Equipped with an IMX219 8MP camera and dual-servo gimbal, the jetson orin nano development kit supports face tracking, object recognition, target tracking, and computer vision projects. Ideal for learning AI vision, edge computing, robotics, and intelligent automation applications
- 11.6-Inch HD Display & AI Voice Assistant: Features an 11.6-inch 1366×768 IPS screen, allowing users to develop and test projects without an external monitor. The built-in AI voice interaction system supports voice commands and intelligent conversations, creating a more engaging and interactive learning experience
- 30 Sensors and 38 Guided Python Tutorials: Features a 30-in-1 sensor board with temperature & humidity, ultrasonic ranging, gas, motion, and other commonly used sensors. Includes 38 guided Python tutorials covering sensor applications, embedded development, and AI visual recognition from beginner to advanced
- Portable All-in-One Design with Rich Expansion Options: The Jetson Orin Nano Dev Kit provides multiple expansion interfaces including I2C/UART/IO interfaces. A custom carrying case integrates all components, making it convenient for classroom teaching, laboratory projects, demonstrations, and mobile AI development
What OpenShell and NemoClaw are intended to do
NVIDIA’s official GTC 2026 coverage identifies two technologies aimed at making agent deployment more controllable:
| Technology | Role described by NVIDIA | What that means for deployment |
|---|---|---|
| OpenShell | Secure runtime and isolation layer | Provides the contained environment in which an agent can run without direct, unrestricted access to the host system. |
| NemoClaw | Policy enforcement, network guardrails and privacy routing | Applies rules to agent behavior, limits network activity and directs data according to privacy requirements. |
NVIDIA says OpenShell and NemoClaw are designed to support safer enterprise deployment and that the technologies could become “the policy engine of all the SaaS companies in the world.” That is NVIDIA’s stated ambition, not an independent assessment of present-day coverage or effectiveness.
Free tools Windows power users keep installed
One-click scans. No signup required.
Neither the event report nor NVIDIA’s GTC material establishes that OpenClaw itself guarantees safe robot control. A robot deployment would still need task-specific permissions, physical emergency stops, testing, monitoring, recovery procedures and a clear human escalation path. Those requirements are especially important when an agent can affect machinery, people or the physical environment.
How OpenClaw fits NVIDIA’s physical-AI strategy
NVIDIA’s GTC 2026 material places OpenClaw inside a larger “physical AI” ecosystem. It lists six Nemotron Coalition model families:
Rank #3
- Supercharged AI Performance: Powered by NVIDIA Jetson Orin NX 16GB, delivers up to 157 TOPS in MAXN Super Mode — ideal for vision AI, robotics, autonomous machines, and generative AI workloads.
- Advanced Thermal Engineering for Full-Power Operation: Equipped with a vacuum copper heat pipe system, ultra-low thermal resistance medium, and high-emissivity black-coated surface combined with high-performance active cooling — ensuring stable full compute power even at 60°C ambient temperature.
- Energy-Efficient & Flexible Power Modes: Adjustable power profile from 10W to 40W, enabling a perfect balance between performance and efficiency for edge AI computing in diverse environments.
- Industrial-Grade Reliability & Design: Ruggedized for operation from -20°C to 60°C at 40W (up to 65°C at 25W), providing dependable performance in industrial automation and outdoor AI deployments.
- Rich Connectivity & AI-Ready Platform: Features 2×RJ45, SIM slot, 4×USB 3.2, HDMI 2.1, CAN, M.2 Key E/M, Mini-PCIe, and 4×CSI camera ports — supporting multi-camera vision, IoT, and robotics projects. Pre-installed with JetPack 6.2 and 128GB NVMe SSD, fully compatible with NVIDIA Isaac, ROS 1/2, and Hugging Face frameworks.
| Model family | Focus identified by NVIDIA |
|---|---|
| Nemotron | Language and reasoning |
| Cosmos | World understanding and vision |
| Isaac GR00T | General-purpose robotics |
| Alpamayo | Autonomous driving |
| BioNeMo | Biology and chemistry |
| Earth-2 | Weather and climate |
The company says it is working with automakers BYD, Hyundai, Nissan and Geely, deploying vehicles with Uber, and collaborating with industrial robotics companies ABB, Universal Robots and KUKA. These partnerships show the breadth of NVIDIA’s physical-AI direction; they do not show that OpenClaw is already integrated into every listed platform.
Where people can experiment with OpenClaw
NVIDIA says attendees at the London event could run a customized OpenClaw agent either on cloud compute or on local accelerated hardware. The concrete local paths it names are an NVIDIA DGX Spark system or a GeForce RTX laptop. NVIDIA also says DGX Spark systems were available for purchase onsite through the NVIDIA Gear Store and Micro Center.
| Deployment path | What the cited material establishes | What it does not establish |
|---|---|---|
| Local accelerated computing | OpenClaw can be run on a DGX Spark or GeForce RTX laptop for local experimentation. | Current retail price, availability by country, benchmark results and the exact model configuration are not stated. |
| Cloud computing | Cloud execution was offered as an alternative at the event. | Provider, region, recurring cost, latency, data-retention terms and performance are not stated. |
NVIDIA separately lists up to 4,000 TOPS of local AI compute and 96 GB of GPU memory for RTX PRO 6000 Blackwell workstation editions. Those figures describe that workstation hardware; they are not an OpenClaw performance guarantee and should not be treated as the specification of a DGX Spark or a particular GeForce RTX laptop.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.The scale behind NVIDIA’s forecast
Huang’s robotics comments were part of a much larger NVIDIA presentation. NVIDIA’s 2026 GTC coverage cites $150 billion invested in venture startups, Huang’s statement that computing demand had increased by one million times over the last few years, and a forecast of at least $1 trillion in revenue from 2025 through 2027. These are NVIDIA’s own figures and projections, not measurements of OpenClaw adoption or robot-safety outcomes.
Rank #4
- 【Core Parameters】★AI Perf: 34/67 TOPS ★GPU:1024-core official Ampere architecture GPU with 32 Tensor Cores ★CPU:6-core Arm Corte-A78AE v8.2 64-bit CPU 1.5MB L2 + 4MB L3 ★Memory:8GB 128-bit LPDDR5 68 GB/s ★Storage: external NVMe via M.2 Key M
- 【Empowered by Large Al Model, Enhanced Human-Computer Interaction】Jetson Orin Super leverages three AI models and incorporates an AI voice interaction module. This multimodal visual system matches the scene being described, enabling environmental awareness and AI visual gameplay. Combined with a large-scale voice module and camera, it enables speech-to-text, semantic analysis, natural conversation, and real-time video analysis, enabling advanced embodied AI applications.
- 【AI Upgrade】Jetson Orin Nano series modules are compact in size but can deliver up to 34-67 TOPS of AI performance, with power consumption ranging from 7 watts to 25 watts. Compared to the Jetson Nano B01, it offers up to 80 times the performance and sets a new standard for entry-level edge AI.
- 【Highly compatible carrier board】Yahboom's carrier board is fully compatible with orin nano module. Compared to carrier boards that use Jetson Nano on the market, the newly upgraded circuit supports 25W power mode, which enables larger and more complex neural networks and fully leverages the performance of the core module. The resources, size, and interfaces of the Yahboom carrier board are consistent with the official board, with the only difference addition of power switch button.
- 【Tutorial materials provided】The JETSON system based on Ubuntu 22.04 provides a complete desktop Linux environment with accelerated graphics, supporting CUDA 12.6, TensorRT 10.7.0, cuDNN 9.6.0, OpenCV 4.10.0, etc. The performance on AI LLM, VLM and visual Transformer is significantly improved compared with the previous generation.
The significance of the OpenClaw discussion is therefore architectural: NVIDIA sees agent software, secure execution environments and physical machines as parts of one computing stack. The safety question is whether permissions, isolation, monitoring and escalation are implemented tightly enough for each real-world task.
Can OpenClaw safely control robots?
Based on the available reporting, the responsible answer is not yet established. Huang described the direction and NVIDIA described proposed runtime and policy components, but neither source supplies independent validation that OpenClaw can safely operate a robot without additional engineering.
A practical evaluation would need to verify, at minimum:
- Which robot functions the agent can invoke and which remain blocked by default.
- Whether permissions are limited to the current task and revoked when the task ends.
- How network access, private data and model prompts are routed and logged.
- How the system detects policy violations and escalates ambiguous commands.
- Whether a human can stop motion immediately through an independent mechanism.
- How the robot behaves after a software, network, sensor or model failure.
Until those controls are demonstrated for a specific robot, OpenClaw should be understood as an agentic-computing platform that may be connected to physical AI—not as a blanket safety certification for autonomous machines.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
