NVIDIA Unveils Jetson Orin Nano 2 to Power the Next Generation of Edge AI and Robotics
NVIDIA has unveiled Jetson Orin Nano 2, a new compact robotics computer aimed at developers building AI-powered machines that need to process information locally and respond to their surroundings in real time.
The new platform is positioned as an entry-level solution for edge AI and physical AI, allowing developers to run increasingly capable AI models directly on devices instead of relying entirely on cloud-based computing.
NVIDIA says Jetson Orin Nano 2 delivers up to 78 trillion operations per second of AI computing, while providing twice the inference performance of the previous Jetson Orin Nano Super. It is also designed to deliver the same performance while consuming significantly less power.
The module and developer kit are expected to become available during the first half of 2027.
Bringing AI Directly Into Machines
The development of AI is increasingly moving beyond computers and smartphones.
Robots, drones, industrial equipment and smart machines are being designed to understand their surroundings, interpret information and make decisions without requiring constant assistance from a remote server.
This requires computing hardware that can operate directly on the device.
Jetson Orin Nano 2 is designed for this purpose, giving developers a platform capable of running AI workloads locally.
For a robot, that could mean analysing camera feeds and recognising objects in real time. For a drone, it could mean processing images and identifying obstacles while flying.
A Significant Performance Upgrade
Jetson Orin Nano 2 includes 8GB of memory and an eight-core Arm CPU, alongside NVIDIA's AI acceleration technology.
The platform features improved Tensor Cores and higher memory bandwidth compared with its predecessor.
NVIDIA says these improvements allow the new system to deliver 2x the inference performance of Jetson Orin Nano Super while maintaining a compact form factor.
For developers, the increased performance could allow more sophisticated AI applications to operate on smaller machines.
Power Efficiency Is Just as Important
AI performance is only one part of the equation for edge devices.
Robots and drones often operate with limited power budgets. A powerful processor that consumes too much energy can reduce operating time or require larger batteries and cooling systems.
NVIDIA says Jetson Orin Nano 2 can deliver the same performance as its predecessor while using 40% less power in a 15-watt operating mode.
That could make the platform particularly useful for mobile robots, autonomous drones and other compact machines.
The Rise of Physical AI
NVIDIA is positioning Jetson Orin Nano 2 around the growing concept of physical AI.
Unlike conventional AI applications that primarily generate content or analyse information, physical AI systems interact with the real world.
A robot might use cameras to identify an object, understand an instruction and then physically move toward that object.
A drone could analyse its surroundings and adjust its flight path.
An industrial machine could recognise components and respond to changes on a production line.
All of these applications require AI processing to happen quickly and reliably.
More Capable Robots
The new platform could help developers build robots capable of handling increasingly complex tasks.
With sufficient AI processing power, a robot can potentially:
Recognise objects and people
Understand natural-language commands
Navigate unfamiliar spaces
Analyse camera and sensor data
Build maps of its environment
Make decisions locally
Respond to changing surroundings
Interact more naturally with humans
The ability to perform these tasks locally is particularly valuable when a robot needs to react within milliseconds.
Drones Are Another Major Target
Autonomous drones are also expected to benefit from more capable edge computing.
A drone operating independently needs to understand its environment while it is moving.
It may need to identify obstacles, interpret visual information and make navigation decisions without waiting for instructions from a remote computer.
NVIDIA says Wing, Alphabet's drone-delivery company, is already using Jetson Orin Nano Super and plans to evaluate Jetson Orin Nano 2 for advancing real-time AI perception and reasoning in its delivery drones.
The new platform could therefore help developers build drones that are more responsive while keeping onboard computing relatively compact.
AI Models Can Run Locally
Another important part of the new platform is its software ecosystem.
Jetson Orin Nano 2 is designed to run modern AI models optimised for efficient inference at the edge.
NVIDIA has highlighted support for models including NVIDIA Cosmos, NVIDIA Nemotron, Gemma 4 and Qwen 3.
This opens the door to applications that combine computer vision with language and multimodal AI.
For example, a robot could potentially interpret a spoken instruction, understand what is visible through its cameras and then determine an appropriate action.
A Large Developer Community
NVIDIA says more than 3 million developers are building on its robotics software ecosystem.
That existing community could give Jetson Orin Nano 2 an advantage as developers begin experimenting with the new hardware.
The platform is also being supported by an ecosystem of hardware companies working on carrier boards, embedded systems and other solutions.
Early Companies Exploring Jetson Orin Nano 2
NVIDIA has identified Cognex, Doosan Bobcat and Matic among companies adopting or exploring the new platform.
Their applications cover different areas, including industrial vision, autonomous equipment and consumer robotics.
The variety of applications illustrates the broad range of potential uses for compact edge AI computers.
Home Robots Could Become Smarter
Consumer robotics is another area where local AI processing could have a major impact.
Home robots operate in environments that constantly change.
People move around, furniture changes position and objects can appear in unexpected locations.
NVIDIA says Matic is using Jetson Orin Nano 2 for home robots with capabilities including conversational AI, gesture recognition, mapping and autonomous cleaning.
More powerful onboard AI could allow household robots to better understand their surroundings and respond to users.
Why Edge AI Matters
Running AI directly on a device offers several potential advantages.
Faster Responses
Local processing reduces the need to send every request to a remote server, helping machines respond more quickly.
Less Dependence on Internet Connectivity
A robot or drone can continue performing important AI tasks even when connectivity is unreliable.
Lower Data Transfer Requirements
Large volumes of camera and sensor information can be processed locally rather than constantly being transmitted to cloud infrastructure.
Greater Privacy
For some applications, keeping information on the device can reduce the need to send sensitive data elsewhere.
Better Support for Mobile Machines
Improved energy efficiency makes advanced AI more practical for battery-powered devices.
A Bigger Shift in AI Hardware
The launch also reflects a wider change in the semiconductor industry.
AI computing was initially dominated by large data centres and powerful accelerator systems.
Now, AI processing is increasingly moving into smaller devices.
This requires chips that can balance performance with energy consumption, physical size and cost.
The Jetson Orin Nano 2 is designed to occupy this space between traditional embedded computing and high-performance AI infrastructure.
The Importance of Local Intelligence
The next generation of robots will require more than mechanical capabilities.
They will need to understand their surroundings.
A modern autonomous machine may need to process video, recognise objects, interpret language, understand its location and decide what to do next.
Putting these capabilities directly onto the machine can make autonomous systems more responsive.
This is one reason edge AI is becoming increasingly important to the robotics industry.
Jetson Orin Nano 2 and the Future of Robotics
The development of smaller and more efficient AI processors could accelerate the adoption of intelligent machines across several industries.
Potential applications include:
Warehouse robotics
Agricultural machines
Delivery robots
Inspection systems
Industrial automation
Autonomous drones
Home robots
Smart cameras
Healthcare robotics
Retail automation
As AI models become more efficient, more powerful capabilities can potentially be deployed on relatively small devices.
Availability
NVIDIA expects the Jetson Orin Nano 2 module and developer kit to become available during the first half of 2027.
The period leading up to launch will give developers and hardware partners time to prepare applications and products around the new platform.
The Bigger Picture
NVIDIA's latest Jetson announcement comes as the AI industry begins moving toward a new stage.
The first wave of generative AI was largely focused on cloud computing, large language models and data-centre infrastructure.
The next wave is increasingly about bringing intelligence into the physical world.
Robots, drones and autonomous machines need to perceive their surroundings, understand information and take action.
That requires AI hardware capable of operating locally.
Jetson Orin Nano 2 is designed to support exactly this transition.
Key Takeaways
Jetson Orin Nano 2 is NVIDIA's latest compact computing platform for edge AI.
It delivers up to 78 trillion AI operations per second.
The platform includes 8GB of memory and an eight-core Arm CPU.
NVIDIA claims 2x the inference performance of Jetson Orin Nano Super.
It can deliver equivalent performance using 40% less power in a 15-watt configuration.
The platform targets robots, drones, vision systems and autonomous machines.
More than 3 million developers are building on NVIDIA's robotics ecosystem.
Cognex, Doosan Bobcat and Matic are among the companies exploring the platform.
Jetson Orin Nano 2 is expected to become available in the first half of 2027.
The Bottom Line
NVIDIA's Jetson Orin Nano 2 is designed to put more powerful AI capabilities directly into the machines that interact with the physical world.
Its combination of improved inference performance, lower power consumption and support for modern AI models could make it an important platform for developers working on robotics, drones and autonomous systems.
The bigger story is the continued shift from AI that lives primarily in the cloud to AI that operates directly inside machines.
As processors become more efficient and AI models become easier to run locally, robots and autonomous devices are gaining the ability to perceive, understand and respond to their environments in real time.
Jetson Orin Nano 2 could become one of the computing platforms helping drive that transition.
