Built on NVIDIA® Jetson Orin™ Nano & Orin™ NX
Each Nexus family pairs a proven compute module with a production carrier, so you scale performance without redesigning your system.

AI mini supercomputer for Jetson Orin Nano and Orin NX. 34 to 157 TOPS on one carrier.
New platforms are on the roadmap. Share your requirements and hear first when they launch.
Get notifiedFully compatible with the Jetson Orin Nano Developer Kit reference carrier. Prototype on the Dev Kit, ship on NEXUS-N1, with no software port in between.
Software, device trees and BSPs written for the Jetson Orin Nano Developer Kit boot as they are.
Every Jetson Orin Nano and Orin NX module fits the same board. Scale without a redesign.
JetPack 6 and JetPack 7 with Jetson Linux, CUDA, TensorRT and DeepStream.
AI performance figures are sparse INT8.
103 x 90.5 x 34.77 mm with module and thermal solution
| Parameter | Orin Nano 4GB | Orin Nano 8GB | Orin NX 8GB | Orin NX 16GB |
|---|---|---|---|---|
| AI performance | 34 TOPS | 67 TOPS | 117 TOPS | 157 TOPS |
| GPU | Ampere, 512 CUDA / 16 tensor | Ampere, 1024 CUDA / 32 tensor | Ampere, 1024 CUDA / 32 tensor | Ampere, 1024 CUDA / 32 tensor |
| CPU | 6-core A78AE, 1.5MB L2 + 4MB L3 | 6-core A78AE, 1.5MB L2 + 4MB L3 | 6-core A78AE, 1.5MB L2 + 4MB L3 | 8-core A78AE, 2MB L2 + 4MB L3 |
| DL accelerator | None | None | 1x NVDLA v2 | 2x NVDLA v2 |
| Memory | 4GB 64-bit LPDDR5, 51 GB/s | 8GB 128-bit LPDDR5, 102 GB/s | 8GB 128-bit LPDDR5, 102 GB/s | 16GB 128-bit LPDDR5, 102 GB/s |
| Storage | SD card, external NVMe | SD card, external NVMe | External NVMe | External NVMe |
| Video encode | 1080p30 on 1 to 2 CPU cores | 1080p30 on 1 to 2 CPU cores | 1x 4K60, 3x 4K30, 6x 1080p60 | 1x 4K60, 3x 4K30, 6x 1080p60 |
| Video decode | 1x 4K60, 2x 4K30, 5x 1080p60 | 1x 4K60, 2x 4K30, 5x 1080p60 | 1x 8K30, 2x 4K60, 4x 4K30 | 1x 8K30, 2x 4K60, 4x 4K30 |
| Power | 7 W to 25 W | 7 W to 25 W | 10 W to 40 W | 10 W to 40 W |
AI performance figures are sparse INT8. Video codec figures are H.265.