Image: iprsoftwaremedia.com
Entry-level edge-AI robotics computer targeting frontier models at lower power.
Key specs
| Performance | 2x inference vs Jetson Orin Nano Super |
|---|
| Power | 40% lower power at same performance (15 W mode) |
|---|
Technical details
Jetson Orin Nano 2 is NVIDIA's new entry-level robotics computer, delivering 78 TOPS of AI compute with 8 GB of memory and an 8-core Arm CPU. NVIDIA claims 2x the inference performance of Jetson Orin Nano Super through improved Tensor Cores and higher memory bandwidth in the same form factor, and 40% lower power for equal performance in 15 W mode. It targets memory-efficient edge inference of small LLMs/VLMs such as Cosmos, Nemotron, Gemma 4 and Qwen 3. Module and developer kit are expected in H1 2027.
| Architecture | Jetson Orin-family edge AI module with improved Tensor Cores and higher memory bandwidth; 8-core Arm CPU. |
|---|
| Params | 78 TOPS AI compute; 8 GB memory |
|---|
| Inputs outputs | Edge robotics computer (camera/sensor input, on-device LLM/VLM/vision inference). |
|---|
| Embodiments supported | Entry-level robots, home robots, delivery and inspection drones, vision AI systems. |
|---|
| Hardware requirements | Same compact form factor as Jetson Orin Nano; 15 W mode cited; module and developer kit expected in the first half of 2027. |
|---|
| Benchmarks claimed | |
|---|
| How to try | developer.nvidia.com/…/jetson |
|---|
| Sources | |
|---|
Benchmarks, where listed, are as claimed by the vendor.
