NVIDIA has announced the Jetson Orin Nano 2, an updated edge AI platform aimed at robotics, drones, machine vision, and other embedded AI applications. The new module provides up to 78 TOPS of AI performance, 8GB of memory, and an 8-core Arm CPU while retaining the compact form factor of the previous Jetson Orin Nano generation.

The Jetson Orin Nano 2 uses an NVIDIA Ampere architecture GPU with updated Tensor Cores and a higher-bandwidth memory subsystem. It includes 8GB of LPDDR5X memory with up to 120GB/s of bandwidth, while the CPU is based on eight Arm Cortex-A78 cores.

According to NVIDIA, the new platform can deliver up to twice the inference performance of the Jetson Orin Nano Super. The company attributes the increase to improvements in Tensor Core performance and memory bandwidth rather than a corresponding doubling of the headline TOPS figure.

Jetson Orin Nano 2 application banner

Power efficiency has also been improved. NVIDIA states that in a 15W operating mode, Jetson Orin Nano 2 can deliver the same level of performance as its predecessor while using about 40% less power. The module supports a power envelope of up to 40W, allowing developers to select operating modes based on workload and thermal requirements.

The platform is intended for applications such as autonomous robots, delivery and inspection drones, and embedded vision systems where AI inference needs to run locally. NVIDIA is also positioning the device for generative AI workloads at the edge, including language and vision-language models.

Jetson Orin Nano 2 works with NVIDIA’s existing Jetson software stack and Jetson agent skills. NVIDIA lists models such as Cosmos, Nemotron, Gemma 4, and Qwen 3 among those that can be used with the platform for memory-efficient edge inference.

NVIDIA logo

Several companies are already evaluating or adopting the new module. Matic plans to use Jetson Orin Nano 2 in home cleaning robots for workloads including mapping, semantic understanding, gesture detection, and conversational interfaces. Alphabet subsidiary Wing, which currently uses Jetson Orin Nano Super in its delivery drones, also plans to evaluate the new platform for onboard perception and reasoning.

Hardware vendors including AAEON, ADLINK, Advantech, Aetina, Antmicro, Auvidea, AVerMedia, Connect Tech, ForeCR, RidgeRun, and Seeed Studio are also preparing carrier boards, complete systems, software, and reference designs for the new module.

Further Information