Kit de développement C100 NVIDIA Jetson Nano 4 Go
215,00 € INCL. VAT
- Jetson Nano 4 GB for learning CUDA, computer vision and AI inference
- 128-core Maxwell GPU, 4 GB LPDDR4 and 16 GB built-in eMMC
- Two CSI camera inputs, 4 USB 3.0, Gigabit Ethernet and 40-pin GPIO
Entreprise, école ou collectivité ? business account Offres éducation
- Description
Description
Jetson Nano 4 GB: the C100 kit to get started with embedded AI and CUDA
The C100 development kit combines the NVIDIA Jetson Nano 4 GB module with a carrier board designed by OKdo and Radxa. It is an affordable gateway to computer vision and neural network inference: Maxwell GPU with 128 CUDA cores, NVIDIA JetPack environment and two MIPI CSI-2 camera connectors.
Its advantage over the classic Jetson Nano kit: the production module includes 16 GB of eMMC, and the board adds a microSD reader to expand the workspace.
Key features of the Jetson Nano C100 kit
- NVIDIA GPU with 128 CUDA cores : GPU-accelerated image classification, object detection or segmentation, with TensorRT and cuDNN provided by JetPack.
- Two MIPI CSI-2 camera inputs : IMX219 and IMX477 sensors are recognized without additional drivers, ideal for stereo or two simultaneous streams.
- 16 GB eMMC + microSD : system on the eMMC or on a larger microSD card for development.
- 4 USB 3.0 ports and Gigabit Ethernet : USB cameras, external storage and fast wired networking.
- M.2 Key E slot : add a Wi-Fi module if the project needs to go cable-free.
- 40-pin GPIO : I2C, I2S, SPI and UART for sensors, motors and displays.
Technical specifications
| Features | C100 Jetson Nano 4 GB kit |
|---|---|
| Module | NVIDIA Jetson Nano 4 GB (production module) |
| Processor | 4-core Arm Cortex-A57 at 1.43 GHz |
| GPU | NVIDIA Maxwell architecture, 128 CUDA cores |
| Memory | 4 GB 64-bit LPDDR4, 25.6 GB/s |
| Storage | 16 GB eMMC 5.1 on the module + microSD reader |
| Video | hardware decoding up to 4K 60 (HEVC), encoding up to 4K 30 |
| Display | Full-size HDMI |
| USB | 4 x USB 3.0 + 1 micro-USB |
| Network | Gigabit Ethernet, M.2 Key E for optional Wi-Fi |
| Cameras | 2 MIPI CSI-2 connectors (IMX219, IMX477 supported) |
| Extension | 40-pin GPIO: I2C, I2S, SPI, UART |
| Power supply | 5 V / 4 A (20 W) via 5.5 x 2.1 mm barrel jack |
| Software | OKdo Ubuntu 18.04 image with NVIDIA JetPack 4.6 |
Which one should you choose?
| Solution | AI compute | Memory | Ideal for |
|---|---|---|---|
| Jetson Nano C100 (this product) | 128-core CUDA Maxwell GPU | 4 GB LPDDR4 | learning CUDA and computer vision, low-budget prototypes |
| Jetson AGX Orin | Ampere GPU with 1792 CUDA cores, over 200 TOPS | 32 GB LPDDR5 | advanced robotics, multi-camera, heavy models |
| Google Coral USB Accelerator | Edge TPU on USB port | that of the host computer | adding TensorFlow Lite inference to an existing board |
Our advice: the Jetson Nano remains the best teaching tool for discovering the CUDA ecosystem. For intensive production use, go for an Orin module. Our comparison Raspberry Pi or NVIDIA Jetson details the differences.
Compatibility and pitfalls to avoid
- Power supply : use a 5 V / 4 A power supply on the 5.5 x 2.1 mm barrel jack. Micro-USB cannot deliver 4 A, and the getting started guide reserves it for the serial console.
- Jumper : do not place a jumper on the 2-pin header next to the barrel jack, contrary to what some tutorials for the original NVIDIA kit say.
- Software : the Jetson Nano stops at the JetPack 4.6 branch (Ubuntu 18.04). Recent JetPack versions target Orin modules. Use the C100 image published by OKdo, not the one for the NVIDIA kit.
- GPIO : the 40-pin pinout matches the Raspberry Pi's, but HATs depend on Raspberry Pi OS drivers: validate each expansion board on a case-by-case basis.
- Cameras : choose IMX219 or IMX477 sensors. Other sensors require a specific driver.
Included and not included
The kit includes the Jetson Nano module mounted on its carrier board. Cables, microSD card and power supply are not included.
Recommended accessories
- Camera Module v2 8 MP : IMX219 sensor natively recognized by the C100 image, to start your vision projects.
- Carte microSD SanDisk Ultra : the getting started guide flashes the system to microSD for more space.
- USB 3.0 SD/TF card reader: to write the c100.img.xz image from your computer.
- HDMI cable: the video output is full-size HDMI.
- RJ45 CAT6 Ethernet cable: Gigabit wired networking for updates and model transfers.
Uses and projects
- Real-time object detection on a camera stream, to count passers-by or monitor an area.
- Mobile robot : vision, line or object following, motor control via GPIO.
- Face recognition with two cameras using Python and OpenCV, the example from the official C100 guide.
- AI teaching : a real CUDA GPU for hands-on labs. See also our article on AI accelerator boards.
Frequently asked questions
What is the difference between the Jetson Nano 4 GB C100 and the NVIDIA kit?
The same Jetson Nano 4 GB module, but in a production version with 16 GB of eMMC, on an OKdo/Radxa carrier board. The software comes from OKdo's C100 image.
Is the Jetson Nano the same as the Jetson Orin Nano Super?
No. The Orin Nano Super is a newer generation, rated at 67 TOPS by NVIDIA. The Jetson Nano uses a Maxwell GPU and the JetPack 4.6 branch.
Which power supply for the Jetson Nano C100?
A 5 V / 4 A (20 W) power supply with a center-positive 5.5 x 2.1 mm barrel jack. Micro-USB is not enough to use full-power mode.
Do I need a microSD card?
The module has 16 GB of eMMC, but the OKdo guide recommends booting from microSD for more development space.
Do Raspberry Pi cameras work?
Yes for IMX219 (camera v2) and IMX477 sensors, supported out of the box. Other models require dedicated drivers.
Can Raspberry Pi HATs be used on the GPIO?
The 40-pin header matches the Raspberry Pi pinout, but the drivers differ. Check each HAT before buying.
Get started with embedded AI with the Jetson Nano 4 GB C100. Shipped within 24/48 h of payment receipt from our stock based in France, official and tested products, returns accepted within 14 days with a return authorization, free pickup point delivery on orders of €200 or more, and payment by Visa, Mastercard, PayPal, Apple Pay or Google Pay. See the full NVIDIA range.
