In this section, you’ll learn what makes the NXP FRDM i.MX 93 board a useful ML development platform, then you’ll boot the board and log in.
The NXP FRDM i.MX 93 board is a practical way to validate ML workloads on real Arm hardware early in your workflow.
It combines:
The following image highlights the i.MX 93 SoC and the NPU location on the board:
Arm Ethos-U65 NPU location
To log in, you connect to the board over a USB-to-UART serial console using picocom. Use a package manager to install it.
brew install picocom
sudo apt update
sudo apt install -y picocom
On macOS, you might need an extra USB-to-UART driver depending on which USB serial bridge the board exposes. You can get the drivers from Silicon Labs Downloads page .
The next step is to establish a USB-to-UART (serial) connection. Connect the board’s debug (DBG) USB-C connector to your computer.
Find the board’s serial device:
If you see more than one candidate device, use this quick method to identify the right one:
ls /dev/tty.*
ls /dev/ttyUSB* /dev/ttyACM* 2>/dev/null
Connect to the board using the picocom command by updating the serial device identifier.
sudo picocom -b 115200 /dev/tty*
You will see the below output on a successful connection.
picocom v3.1
...
Terminal ready
Now you’re ready to log in to Linux on the board. Connect the board’s POWER USB-C connector to your laptop and toggle the OFF/ON switch if it’s set to off. You should see one red and one white light on the board. In the picocom terminal window, the board will print boot logs. The last line should end with a login prompt.
...
[ OK ] Reached target Graphical Interface.
Starting Record Runlevel Change in UTMP...
[ OK ] Finished Record Runlevel Change in UTMP.
NXP i.MX Release Distro 6.6-scarthgap imx93frdm ttyLP0
imx93frdm login:
Type root to log in as root (no password is needed).
If you connect the board to a display, you can see ML demos and application output directly on the device. This is useful when you want quick visual confirmation that your model runs on the target hardware. As per the setup instructions in NXP’s getting started guide :
The demos will then show up automatically.
NXP board built-in ML demos