Overview
The NVIDIA Jetson Nano 4GB module brings CUDA-accelerated AI to cost-sensitive devices. Its 128-core NVIDIA Maxwell GPU delivers 472 GFLOPS, enough to run classification, detection and segmentation networks alongside multiple 1080p video streams, while consuming between 5W and 10W.
Unlike the developer kit, the production module includes 16GB of eMMC storage and uses the 260-pin SO-DIMM connector shared with Jetson TX2 NX, Xavier NX and Orin Nano, so it fits a wide ecosystem of carriers and enclosures.
Key Features
- 128-core Maxwell GPU: 472 GFLOPS of FP16 compute
- Quad-core Cortex-A57: Arm cores at 1.43 GHz
- 4GB LPDDR4: 64-bit interface, 25.6 GB/s
- 16GB eMMC: On-module storage for production units
- 5W to 10W: Two power modes for fanless designs
- SO-DIMM form factor: 69.6 x 45 mm, 260-pin, broad carrier ecosystem
Technical Specifications
| Specification | Details |
|---|---|
| GPU | 128-core NVIDIA Maxwell GPU |
| CPU | Quad-core Arm Cortex-A57, 1.43 GHz |
| Memory | 4GB 64-bit LPDDR4, 25.6 GB/s |
| Storage | 16GB eMMC 5.1 |
| Video Encode | 4K30, 4x 1080p30, 9x 720p30 (H.264/H.265) |
| Video Decode | 4K60, 2x 4K30, 8x 1080p30, 18x 720p30 (H.264/H.265) |
| Camera | 12 lanes MIPI CSI-2 (up to 4 cameras) |
| PCIe | 1x1 + 1x4 PCIe Gen2 |
| Networking | 10/100/1000BASE-T Ethernet |
| Display | HDMI 2.0 and eDP 1.4 |
| USB | 1x USB 3.0, 3x USB 2.0 |
| Power | 5W to 10W |
| Form Factor | 69.6 mm x 45 mm, 260-pin SO-DIMM |
Ideal Use Cases
- Entry-level smart cameras and kiosks
- Educational and maker projects moving to production
- Low-power IoT gateways with on-device inference
- Spare modules for deployed Jetson Nano products
Why Choose This Product?
Jetson Nano remains the lowest-cost way to run the NVIDIA software stack on a production module. For new designs that need more performance, Jetson Orin Nano fits the same connector.
Interested? Contact us for pricing, volume orders and deployment support.
