NVIDIA® Jetson Nano™ Developer Kit - B01

I'm pretty happy so far. Setting up python enviro nment a little finicky - suggest you use virtualenv.
Don't try 16GB, you will run out of space quickly. Get a 160MB/s microSD card at least 32GB. I created a 4GB swapfile which improved dev env build speed

I'm pretty happy so far. Setting up python enviro nment a little finicky - suggest you use virtualenv.
Don't try 16GB, you will run out of space quickly. Get a 160MB/s microSD card at least 32GB. I created a 4GB swapfile which improved dev env build speed

Nice quick quad core arm64 SBC with a decent amount of computing power available from the nVidia GPU.
It is definitely a specialised board. While it can be used for watching videos, or running OpenGL or GLES games it's really not for that.

It takes a while to load what it needs to utilise CUDA as a compute device, so the overhead on short tests is large. Once it is loaded it's quite fast, so it's good in real world application.

One thing to be wary of is the OS image is set to full speed 10 watt mode, but the hardware is configured to use MicroUSB for power ie 5W mode. Either set the OS to lower power mode or use a jumper (not included) to set the board to use a higher wattage power supply via the barrel connector.
If either of these is not done, the board will crash under load because of insufficient available curre

This is hands down the fastest ARM-based device ever used with Ubuntu running on it.

It has to use a 4amp power support like https://www.amazon.com/gp/product/B01N4HYWAM

And also have to set the power jumper to use that jack with these
https://www.amazon.com/gp/product/B00N552DWK

Llegué temprano y estaba completamente sellado. Debido a la potencia de los gráficos, se calentará un poco, por lo que requiere un ventilador