STMicroelectronics STM32N6570-DK
Last updated
Was this helpful?
Last updated
Was this helpful?
The STM32N6570-DK Discovery kit is a development board with the STM32N657X0H3Q Arm® Cortex®‑M55‑based microcontroller featuring ST Neural-ART Accelerator, H264 encoder, NeoChrom 2.5D GPU and 4.2 MB of contiguous SRAM. The kit also includes a camera and microphone with extensions to support other sensors. The kit is fully supported by Edge Impulse. You'll be able to sample raw data, build models, and deploy trained machine learning models directly from the studio.
To set this device up in Edge Impulse, you will need to install the following software:
STM32 Cube Programmer. The firmware provided by Edge Impulse uses the CLI abilities of this program. Please add the program's location to your PATH using your operating system instructions.
Use a USB-C cable to connect the development board to your computer to the USB-C connector CN6 marked with STLINK-V3EC. Depending on your connection you may need additional power which can be added via the USB-C connection CN8 marked USB1.
If you are powering the board with just CN6 then place jumper JP2 on the 2 pins on the top (1-2:5V_STLK).
If you are powering the board with CN6 and CN8 then place jumper JP2 on the 2 pins in the middle (3-4:5V_USB_STLK).
If you have a new kit you may need to update the firmware of the onboard ST-LINK. Open the STM32 Cube Programmer application and follow the instructions found here on ST's website. Typically, you only need to click a few buttons to easily update the firmware on the ST-LINK device.
Three binaries must be programmed in the board external flash using the following procedure:
Download the default Edge Impulse firmware, model, and bootloader here. See the Readme and the scripts for the appropriate commands to flash the three files.
Switch BOOT1 on the board switch to right position and reset the board using button B1 labeled NRST
Program ai_fsbl_cut_2_0.hex (First stage bootloader)
Program network_data.hex (params of the networks; To be changed only when the network is changed)
Program firmware-st-stm32n6.bin (firmware application)
Switch BOOT1 on the board to left position and reset the board using button B1 labeled NRST
When deploying a new binary both the network_data.hex and the firmware-st-stm32n6.bin need to be flashed. The bootloader only needs to be programmed once.
Please visit the STM32N6 Series site for further programming information
To start acquiring data from the device open a command prompt or terminal, run:
This will start a wizard which will ask you to log in, choose an Edge Impulse project, and establish communication between your development kit and the Edge Impulse project chosen. If you want to switch projects run the command with --clean.
At this time only Object Detection projects using FOMO or YOLOV5 are support with the N6 on Edge Impulse.
To start inferencing with a model on this device you need to complete an Edge Impulse project and download the binary from the project and flash using the above steps. Once flashed you may use the command:
That's all! Your device is now connected to Edge Impulse. To verify this, go to your Edge Impulse project, and click Devices. The device will be listed here.
With everything set up you can now build your first machine learning model with these tutorials:
There are two deployment options available for the STM32N6. If you are using the dev kit with the default connections you may get a fully built binary for your kit using the ST STM32N6 option. If you are wanting to work with source code and use on your own device please select the ST Neural-ART library deployment option. The ST Neural-ART option will generate source code that will use the ST accelerator found on the N6 series.
Errors during flashing:
also
Fully remove power from the board and then reconnect power to the board. Do not rely on the reset button to clear this error.
USB C Cables: Inspect USB cables and swap with ones that have both power and data lines. Attempt again with certified USB cables.
