NVIDIA has unveiled the Jetson Orin Nano 2, a new entry-level edge AI robotics computer designed to open up physical AI development — from vision AI platforms to autonomous delivery drones and monitoring devices — to millions of engineers worldwide. According to Deepu Talla, Vice President of Robotics and Edge AI at NVIDIA, the platform brings frontier-class AI performance to low-power, affordable edge hardware for the first time.
From Data Centre to Desktop-Sized Edge Device
During a pre-briefing, Deepu Talla explained that today's small and medium frontier models have reached the accuracy of last year's largest frontier models, unlocking real-time intelligence for edge devices. He noted that state-of-the-art models once restricted to cloud data centres can now execute instantaneously on basic Jetson platforms.
Talla also emphasised that the Jetson Orin Nano 2 allows autonomous machines and edge hardware to process top-tier vision and language models locally and without delay, unlocking capabilities that were previously out of reach for entry-level systems.
AI Processing Power
NVIDIA claims the economical, low-power platform sets a new benchmark for entry-level edge processing, bringing high-end generative AI capabilities to a massive software community. The system packs an 8-core Arm processor, 8GB of system memory and 78 TOPS of total AI computing power.
Equipped with enhanced Tensor Cores and expanded memory bandwidth, the Jetson Orin Nano 2 doubles the inference throughput of the Jetson Orin Nano Super while occupying the exact same footprint. Operating at just 15 watts, it matches previous-generation performance levels using 40% less energy.
The new platform allows developers to run large language models and vision-language systems directly on the hardware, with support for lightweight open architectures including Gemma 4, Qwen 3, NVIDIA Nemotron and NVIDIA Cosmos.
"The Jetson Orin Nano 2 computer puts that breakthrough within reach of millions of developers, delivering the performance and energy efficiency needed for real-time reasoning at the edge."— Deepu Talla, Vice President of Robotics and Edge AI, NVIDIA
Key Strategic Ecosystem Partners
A broad ecosystem of hardware and software partners — including AAEON, Advantech, Aptiv, Connect Tech and Seeed Studio — is already designing compatible carrier boards, specialised computing systems, custom AI software and turnkey reference blueprints to speed up commercial rollout for clients.
AAEON develops rugged carrier boards and compact system-level hardware tailored specifically for the NVIDIA Jetson architecture, enabling enterprise teams to move edge AI algorithms from desktop prototypes into harsh, real-world deployment environments.
Advantech builds purpose-built carrier boards, thermal management systems and turnkey hardware configurations for the Jetson ecosystem, helping developers move vision-language and autonomous machinery projects into mass commercial rollout.
Aptiv integrates technologies like its PULSE 360-degree sensing with NVIDIA Jetson compute, helping customers bridge the gap between edge AI proof-of-concepts and scalable, automotive-grade deployments across autonomous mobile robots, inspection drones and commercial equipment.
Connect Tech provides high-performance carrier hardware, custom thermal solutions and board support packages for Jetson modules, allowing teams to right-size edge compute for tight space, power and environmental constraints.
Seeed Studio offers carrier boards, modular enclosures and custom AI software stacks, combining Jetson compute with plug-and-play hardware kits to accelerate rollout for developers building smart agriculture tools, vision AI inspection systems and autonomous delivery units.