NVIDIA Jetson Orin Nano Establishes a New Standard for Entry-Level Edge AI and Robotics

 With an 80x performance leap, NVIDIA Jetson Orin Nano Establishes a New Standard for Entry-Level Edge AI and Robotics

GTC — With the introduction of the Jetson Orin NanoTM system-on-modules, NVIDIA has expanded the NVIDIA® JetsonTM lineup and raised the bar for entry-level edge AI and robotics. These new products can perform up to 80 times faster than the previous generation.

Within six months of its release, Canon, John Deere, Microsoft Azure, Teradyne, and TK Elevator joined over 1,000 other customers who have adopted the Jetson Orin family.

The NVIDIA Jetson family now covers six Orin-based production modules for the first time to serve the complete spectrum of edge AI and robotics applications. This ranges from the Orin Nano, which offers the fastest Jetson compact factor and up to 40 trillion operations per second (TOPS) of AI capability, to the AGX OrinTM, which offers 275 TOPS for very sophisticated autonomous machines.

An NVIDIA Ampere architecture GPU, Arm-based CPUs, cutting-edge deep learning and vision accelerators, high-speed interfaces, rapid memory bandwidth, and support for multimodal sensors are all included in the Jetson Orin. More clients are able to commercialise products thanks to this performance and adaptability, including engineers implementing cutting-edge AI applications and developers creating the next generation of intelligent robots using the Robotics Operating System (ROS).

According to Deepu Talla, vice president of embedded and edge computing at NVIDIA, "over 1,000 customers and 150 partners have adopted Jetson AGX Orin since NVIDIA announced its availability just six months ago, and Orin Nano will considerably increase this acceptance." "Jetson Orin is the right platform for nearly any sort of robotics deployment imaginable," according to the company. "Millions of edge AI and ROS developers currently benefit from orders-of-magnitude increases in performance."

Making Accessible Cutting-Edge AI and Robotics

The Orin Nano modules are pin- and form-compatible with the Orin NX modules, which were previously revealed. Customers may begin creating applications for the Orin Nano series right away with the AGX Orin development kit thanks to full emulation compatibility. Customers now have the freedom to create a single system to accommodate several Jetson modules and simply expand their applications.

Orin Nano's high-speed I/O and NVIDIA Ampere architecture GPU allow several concurrent AI application pipelines. Developers of low-cost, more complicated AI models gain from easy access to higher-level models when creating entry-level devices and applications like retail analytics and industrial quality control.

Two variations of the Orin Nano modules will be offered. The 4GB version of the Orin Nano gives up to 20 TOPS with power choices as low as 5W to 10W, while the 8GB model delivers up to 40 TOPS with power customizable from 7W to 15W.

Over 700,000 ROS developers now have access to accelerated processing thanks to the Jetson Orin platform, which was created to tackle the most difficult robotics problems. Roboticists may now achieve higher speed and productivity thanks to updates in the most recent NVIDIA IsaacTM software for ROS and the robust hardware capabilities of Orin Nano.

Support for the ecosystem and software is strong.

Wide-ranging support for Jetson Orin has come from companies including Canon, John Deere, Microsoft Azure, Teradyne, TK Elevator, and many more in the robotics and embedded computing ecosystem.

Two variations of the Orin Nano modules will be offered. The 4GB version of the Orin Nano gives up to 20 TOPS with power choices as low as 5W to 10W, while the 8GB model delivers up to 40 TOPS with power customizable from 7W to 15W.

Over 700,000 ROS developers now have access to accelerated processing thanks to the Jetson Orin platform, which was created to tackle the most difficult robotics problems. Roboticists may now achieve higher speed and productivity thanks to updates in the most recent NVIDIA IsaacTM software for ROS and the robust hardware capabilities of Orin Nano.

Support for the ecosystem and software is strong.

Wide-ranging support for Jetson Orin has come from companies including Canon, John Deere, Microsoft Azure, Teradyne, TK Elevator, and many more in the robotics and embedded computing ecosystem.

With more than one million developers, 6,000 clients, including 2,000 startups, and 150 partners, the NVIDIA Jetson ecosystem is expanding quickly. Jetson partners provide a wide spectrum of assistance, including services for designing AI hardware, software, and applications as well as cameras, sensors, and other peripherals, developer tools, and development systems.

Orin Nano is powered by the same NVIDIA CUDA-XTM accelerated computing stack used to produce ground-breaking AI solutions in industries including industrial IoT, manufacturing, smart cities, and more. It is supported by the NVIDIA JetPackTM software development kit.

The Jetson Orin Nano modules will be offered to start in January for $199.