NVIDIA continues to push the envelope of AI accelerators – both in the data center and at the edge. Last month, it announced the availability of the Jetson Orin Nano Developer Kit, the latest addition to the Jetson family of devices.
Initially announced in September 2022, the Jetson Orin Nano system-on-module (SoM) delivers 80x the performance of the previous generation Jetson Nano device. The developer kit puts the power of the SoM in the hands of developers by making it accessible.
Below is a snapshot of the benchmark that compares the performance of AI vision models on Jetson Nano and Jetson Orin Nano.
Jetson Orin has a graphics processing unit (GPU) based on the NVIDIA Ampere architecture, an ARM CPU, next-generation accelerators for deep learning and vision, high-speed interfaces, fast memory bandwidth, and support for multiple types of sensors.
The performance and flexibility enable customers to build products with in-built, advanced intelligence that seemed impossible before. With this device, developers can infuse AI into applications running at the edge, while makers and builders can leverage Robotics Operating System (ROS) to build intelligent machines.
The Jetson Orin Nano 8 GB module has an NVIDIA Ampere architecture GPU with 1024 CUDA cores, 32 third-generation Tensor Cores, and a 6-core ARM CPU.