- Last Update: Dec 23, 2024. Promoted a book "A Quick Guide to Coding with AI" with affiliate links.
NVIDIA Jetson Orin Nano:
- Unboxing: Comes with a robust power supply, quick start manual (kinda), and the device itself. Needs an SD card purchased separately. Has various ports like HDMI, USB, Ethernet, GPIO, and video input.
- Specs: Has a six-core ARM Cortex 64-bit CPU, NVIDIA Ampere architecture GPU, 8 GB RAM, Ethernet and wireless connectivity, and NVMe support. Considered a supercharged Raspberry Pi for AI.
- Setup: Firmware update to version 36.x via temporarily installing Jetpack 5.1.3, followed by installing Jetpack 6.1.
- Performance Tests: With Ollama, it handled 1 billion parameter models easily, 3 billion with slightly longer loading times, but struggled with 7 billion parameter models.
- Pros: Compact, lightweight, impressive performance for the price, quiet operation, and user-friendly OS.
- Cons: Limited to handling models up to 7 billion parameters comfortably and occasional glitches like random lockups and "system throttled due to overcurrent" error.
- Final Verdict: Worth it for developers exploring AI at the edge or prototyping. Not for heavy-duty production workloads but offers great value.
- Where to Order: Currently out of stock but can be ordered from Amazon, Arrow Electronics, Sparkfun, and Seeed Studio.
- Encouraged to explore more with the Orin Nano in upcoming projects, ask questions or leave comments on social media.
**粗体** _斜体_ [链接](http://example.com) `代码` - 列表 > 引用
。你还可以使用@
来通知其他用户。