On August 25, 2026, NVIDIA announced the release of NVIDIA Jetson Orin Nano 2, targeting robotics and entry-level edge AI applications. This blog will explain what Jetson Orin Nano 2 is, highlight its key specifications and performance improvements, and examine where it fits within the NVIDIA Jetson portfolio for robotics, machine vision, and autonomous systems.

What Is NVIDIA Jetson Orin Nano 2? 

NVIDIA Jetson Orin Nano 2 is a compact edge AI computing platform designed for entry-level robotics, machine vision, autonomous systems, and generative AI workloads. 

It extends the Jetson Orin Nano family with higher AI performance, more CPU cores, faster memory, and improved performance per watt while maintaining the same compact module form factor. NVIDIA positions Jetson Orin Nano 2 as an entry-level platform for bringing more capable AI models closer to where data is generated.

NVIDIA Jetson Orin Nano 2 Key Specifications

• 78 TOPS INT8 AI performance
• NVIDIA Ampere GPU with 1,536 CUDA cores
• 8-core Arm Cortex-A78 CPU
• 8GB LPDDR5X memory
• 120 GB/s memory bandwidth
• PCIe Gen4 expansion
• 15W to 40W power range

These specifications give Jetson Orin Nano 2 more compute and memory bandwidth than the previous Jetson Orin Nano 8GB while keeping the same physical module dimensions.

What Has Improved in Jetson Orin Nano 2?

The biggest change is not just the increase in TOPS. Jetson Orin Nano 2 also brings more GPU resources, additional CPU cores, and faster memory.

The previous Jetson Orin Nano 8GB reaches 67 TOPS, while Jetson Orin Nano 2 increases that figure to 78 TOPS. GPU resources increase from 1,024 CUDA cores to 1,536 CUDA cores, while the CPU moves from 6 to 8 Cortex-A78 cores. Memory bandwidth also rises from 102.4 GB/s to 120 GB/s.

Those improvements matter because real edge AI workloads rarely depend on a single specification. A robot may be processing camera feeds, running perception models, handling navigation logic, and communicating with other sensors at the same time. More CPU resources and higher memory bandwidth can help keep those workloads moving alongside GPU inference. 

NVIDIA also positions Jetson Orin Nano 2 as delivering up to a 2X performance leap compared with the previous entry-level platform, depending on the workload. 

Better Performance with Less Power

Power is just as important as compute in robotics. 

Mobile robots, drones, and compact autonomous systems are often limited by battery capacity, enclosure size, and thermal design. Adding more AI performance is useful only if the system can support the additional power and heat. 

Jetson Orin Nano 2 is designed to scale from 15W to 40W. NVIDIA also highlights that the platform can achieve comparable performance to Orin Nano Super at 15W versus 25W, representing about 40% lower power consumption at the same performance level.

That gives system designers more flexibility. A robotics platform can operate at a lower power point when efficiency matters most, or scale higher when the workload needs additional compute headroom.

Where Does It Fit in the Jetson Family?

NVIDIA continues to scale Jetson across different levels of edge AI performance. 

Jetson Orin Nano 2 sits at the entry level, while Jetson Orin NX addresses more demanding mainstream workloads and higher-end Jetson platforms target advanced robotics and physical AI applications.

The key question is therefore not simply whether Jetson Orin Nano 2 is faster than the previous Nano platform. It is whether its combination of performance, power, and form factor is enough for the intended application. 

For many compact robotics and vision systems, that balance may be more important than moving directly to the highest-performance Jetson module available.

Target Applications for NVIDIA Jetson Orin Nano 2

According to NVIDIA, Jetson Orin Nano 2 is designed for entry-level robotics and edge AI applications, including:

• Robots — supporting real-time perception, navigation, interaction, and reasoning 
• Delivery drones — enabling onboard AI perception and autonomous operation 
• Inspection drones — processing vision and AI workloads directly at the edge 
• Vision AI systems — supporting real-time visual perception and intelligent analysis 
• Generative AI and physical AI — running language and vision-language models on compact edge devices

These applications take advantage of Jetson Orin Nano 2’s higher AI performance and improved power efficiency to bring more real-time intelligence directly onto edge devices.

Jetson Orin Nano 2 Availability

NVIDIA Jetson Orin Nano 2 is expected to be available in the first half of 2027. With the module maintaining the same compact mechanical form factor as the existing Jetson Orin Nano, it also provides a path for edge AI systems already designed around the Jetson Orin ecosystem to adopt the new platform. 

This includes Premio’s JCO-1000-ORN and JCO-3000-ORN Series, which currently support NVIDIA Jetson Orin Nano and Orin NX modules for rugged edge AI deployments. Premio plans to incorporate NVIDIA Jetson Orin Nano 2 into both series, extending the platforms with the increased AI performance and power efficiency of the new module.

jco-1000-orn jco-3000-orn