NVIDIA Announces Jetson Orin Nano 2 with Double AI Performance or 40% Energy Savings
NVIDIA has announced the Jetson Orin Nano 2, a next-generation edge AI computer for robotics slated for release in the first half of 2027. The new module delivers twice the AI inference performance of its predecessor, or a 40% reduction in energy consumption when matching the previous generation's performance. This hardware upgrade enables real-time execution of modern small and medium-sized AI models directly on compact edge devices. It will significantly benefit developers in robotics and embedded systems by providing higher efficiency and processing power without increasing physical size. The Jetson Orin Nano 2 features 78 TOPS of AI compute, an 8-core Arm CPU, 8GB of memory, and shares the same form factor as the Jetson Orin Nano Super. When configured to match the performance of the previous generation, it consumes only 15W of power.
## BACKGROUND
NVIDIA's Jetson family consists of low-power embedded computing boards designed to accelerate machine learning applications directly on local devices, a practice known as Edge AI. Running AI at the edge reduces latency and bandwidth usage compared to cloud-based processing. TOPS (Tera Operations Per Second) is a standard metric used to quantify the maximum computing performance of AI hardware.