← All hardware
The most affordable on-ramp to generative edge AI, with 67 TOPS for $249.
Pros
- Best-in-class software ecosystem (CUDA, TensorRT, Isaac, Metropolis)
- Exceptional price/performance after the Super update
- Free performance boost delivered to existing owners via software
Cons
- 8GB memory limits larger generative models
- Power draw (7-25W) higher than pure NPU accelerators
- Requires separate NVMe and storage for serious work
✓ Where it shines / best for
- Affordable entry into edge AI and generative-AI prototyping
- Students, makers, and startups building vision and robotics projects
- Running small LLMs and vision transformers on a budget
✕ Not the best fit for
- Large LLMs or heavy multimodal models (limited 8GB memory)
- High-throughput multi-camera industrial deployments needing AGX-class compute
- Workloads requiring more than ~67 TOPS
Features
- ✓ Up to 67 INT8 TOPS - a 1.7x boost over the original Orin Nano via the 'Super' software update
- ✓ 1024-core NVIDIA Ampere GPU with 32 Tensor Cores
- ✓ 6-core Arm Cortex-A78AE CPU at 1.7 GHz
- ✓ 8GB LPDDR5 memory with 102 GB/s bandwidth
- ✓ Configurable power modes from 7W to 25W
- ✓ Full JetPack SDK: CUDA, TensorRT, plus tuned LLM, VLM, and vision-transformer examples
- ✓ Carrier board accepts all Orin Nano and Orin NX modules; M.2, CSI, USB, GPIO
Pricing
| Plan | Price | Billing | Notes |
|---|---|---|---|
| Jetson Orin Nano Super Developer Kit | $249 | one-time | Price cut announced Dec 2024 with the 'Super' update; existing Orin Nano owners get the boost free via JetPack |
Pricing verified from the official source. Prices change often — confirm on the vendor's site before buying.
Specifications
| power | 7-25W |
| memory | 8GB LPDDR5, 102 GB/s |
| form_factor | Developer kit with carrier board |
| architecture | NVIDIA Ampere GPU (1024 CUDA cores, 32 Tensor cores) + 6-core Arm Cortex-A78AE |
| ai_performance | 67 INT8 TOPS |
Sponsored
A full review is being generated for this product and will appear here shortly.