nvidia 1 in stock
NVIDIA Jetson edge-AI hardware in Manila. Authentic developer kits, modules, and accessories for makers building robotics, computer vision, and on-device generative AI. Power-of-a-workstation in a Raspberry-Pi-sized footprint. Same-day shipping from Quezon City, 4 PM cutoff weekdays.
featured kit
Jetson Orin Nano Super Developer Kit — 67 TOPS AI performance, 1024-core Ampere GPU with 32 Tensor cores, 6-core Cortex-A78AE CPU, 8 GB LPDDR5 at 102 GB/s. Runs JetPack 6 and Ubuntu 22.04 out of the box. Designed for local LLM inference, multi-camera robotics, and on-device vision pipelines.
specs at a glance
| AI performance | up to 67 TOPS (sparse INT8) |
|---|---|
| GPU | 1024-core NVIDIA Ampere + 32 Tensor cores |
| CPU | 6-core Arm Cortex-A78AE @ 1.7 GHz |
| Memory | 8 GB 128-bit LPDDR5 @ 102 GB/s |
| Storage | microSD + M.2 Key M (NVMe SSD) |
| Power | 7–25 W configurable |
| I/O | 4× USB 3.2, GbE, DisplayPort, 40-pin GPIO, MIPI CSI |
| OS | JetPack 6 / Ubuntu 22.04 LTS |
common uses
- On-device LLMs — run Llama 3 8B, Phi-3, Mistral, and Gemma quantised models locally without cloud round-trip
- Robotics + ROS 2 — multi-camera SLAM, depth perception, manipulation pipelines
- Computer vision — YOLOv8 / YOLOv11 at 30+ FPS, multi-stream RTSP analytics
- Smart camera + retail analytics — people counting, license-plate recognition, queue monitoring
- Generative AI demos — Stable Diffusion, Whisper STT, edge voice agents
- Drones + AGVs — visual-inertial odometry, obstacle avoidance, autonomous navigation
notes
Jetson kits are 5V-tolerant on the 40-pin GPIO but use 3.3V logic — same as Raspberry Pi. Order an NVMe SSD if you plan to run large models — microSD throughput bottlenecks LLM token rates. Active cooling is included on the Super dev kit; the original Orin Nano needs an aftermarket fan. Ships from Manila with same-day cutoff at 4 PM weekdays.
FAQ
Is this the original Orin Nano or the new Super?
This is the Super developer kit — the 2024 firmware-unlocked version with 67 TOPS (vs 40 TOPS on the original) and 102 GB/s memory bandwidth (vs 68 GB/s). Same hardware, new power profile.
Can I run Llama 3 or Mistral locally on this?
Yes. With 8 GB LPDDR5 and 67 TOPS, you can run Llama 3 8B (Q4), Phi-3 Mini, and Gemma 2B at usable token rates. Use NVIDIA's `jetson-containers` repo or Ollama with CUDA backend.
Does it work with Raspberry Pi camera modules?
The 2× MIPI CSI-2 lanes accept IMX219 (Pi Cam v2) and IMX477 (HQ Cam) directly. Pi Camera v3 (IMX708) is not officially supported but works with community drivers.
What OS does it ship with?
JetPack 6 (Ubuntu 22.04 LTS + CUDA 12 + cuDNN + TensorRT). Flash via NVIDIA SDK Manager or pre-imaged microSD from our store.
Power supply included?
Yes — official NVIDIA 19V/4.74A barrel-jack PSU ships with the developer kit. Use a quality NVMe SSD if running models > 4 GB.
Warranty + returns?
1-year NVIDIA manufacturer warranty against defects. 7-day local inspection window — if DOA, email us with serial and we replace.
// filters
1 left