NVIDIA Jetson Nano 2GB Developer Kit available now

We are excited to introduce the Jetson Nano 2GB Developer Kit, available now for $59!


SD card…https://developer.nvidia.com/jetson-nano-2gb-sd-card-image

The Jetson Nano 2GB Developer Kit is a great way to get started with hands-on AI, deep learning, and computer vision. We have also launched the Jetson AI Fundamentals course and certification program for learning the ins-and-outs of AI.

The new kit is shipping and available to order now.


Order placed. Looking forward to working with it later this month. :)

Is that USB-C I see? YESS!!

Also, I see the fan header and AC wi-fi adapter say “Not initially available in all regions” - can you clarify which regions those are? Will the US version have them?

Yes, the US/Canada version has both. In other regions, the footprint for the fan header is there, in initial shipments it would just need the 4-pin header soldered on. And the 802.11ac WiFi dongle will be added in other regions as it becomes available.

1 Like

Would it have 32 or 64 bit L4T?

64-bit (same as original Nano). It has been some time since a 32-bit Ubuntu was shipped with L4T.

I preordered today, I’m going to miss my TK1 but at the same time i’m ecstatic to be able to try my ideas and creations out on this new board, my only worry about ram limitations is for tiny_yolo segmentation faults, i’m hoping the 2gb nano will somehow be performing magic.
I just finished a tiny omni wheel bot too.
Oh, and it’s my birthday on the 23rd if you can make this date I won’t mind at all :P


This is amazing embedded computing powerhorse for 59$… I already love it.

BTW, do you have any benchmark with this platform of popular inference TRT models such as the ones available from jetson-inference ?

Hi @Honey_Patouceul, there are DNN inferencing benchmarks included in the blog here: https://developer.nvidia.com/blog/ultimate-starter-ai-computer-jetson-nano-2gb-developer-kit/

Also in my testing with jetson-inference, once loaded the runtime performance is the same as Nano 4GB (i.e. 22FPS with SSD-Mobilenet-v2-COCO). All of the jetson-inference models run on Nano 2GB.

The screencast videos I filmed for the Jetson AI Fundamentals course were shot on the Nano 2GB (see this new YouTube playlist starting at S3E1)

1 Like

What is the HDMI version of the display output?

How much memory is available with 2G mem?


This looks fantastic, NVIDIA, thanks!

I’m wondering about the camera. So this device has TWO mipi lanes exposed, right?

Does the underlying Nano SOM still have 12 mipi lanes, of which only 2 are connected to the carrier board?

Hi nono, it is HDMI 2.0.

Hi @lthbk2959, from my notes, there is ~1967MB available for userspace OS. If you have display attached and are using LXDE, the memory usage after boot is around 500MB.

Hi @360cities, yes the MIPI CSI connector is 2 lanes.

The Nano 2GB module is for devkit use only and not intended for production, so that is unspecified.

May I ask:

  1. is there power solder/header pins available on the board in case I want to power using battery?
  2. Are there other Linux distros available other than Ubuntu? I used to use Puppy or Damn Small Linux back in the day and i’m thinking about similar directions after flashing the life from my TK1. I would be happy to install my own packages.
  3. Audio options other than HDMI - are there composite out pins? or recommended USB audio devices?
    thanks for keeping us all sane in 2020

Hi, does the Jetson Nano 2GB fit into cases/enclosures which have been designed for the 4GB model? If not, are there already any 2GB model compatible cases available?
Thx in advance!

Hi i6mods, similar to the 40-pin header on the Nano 4GB, the Nano 2GB has two 5V pins on it’s 40-pin header that can accept up to 2.5A each to power the devkit.

They have been some ports done by the community to other Linux distributions, for example Yocto and Arch Linux - for more info see this thread:


Note that these were done for Nano 4GB, but should be similar for Nano 2GB.

Like the Nano 4GB, there is also I2S on the Nano 2GB’s 40-pin header. Regarding USB audio devices, if it is a standard audio device with driver support in Linux kernel 4.9, it should work. For example I use my Logitech headset on it and various USB mics that don’t require special driver support.

The mechanical size of the kit is the same, and the mounting holes are in the same locations. The footprint of the fan header is in the same location. The differences would mainly be in the ports - Nano 2GB has USB-C power instead of DC barrel jack, 3 USB ports, and HDMI (no DisplayPort). So you might check if the cut-out of the enclosure can still match the ports. I believe many of the enclosure/case designers are checking against this and will update their designs as needed.

1 Like



Currently, the development and mass production process is being organized with jetson nano.

  1. Can you provide a table of the difference in performance and price between jetson nano(b01) and jetson nano 2GB?

  2. And I am curious if the jetson nano 2G module (for production) will be released at the same time as the jetson nano 2G devkit module.

Thank you.