Hi,
I am working on STM32N6570-DK Edge AI development on Windows and I am facing a board programming error when using stm32ai-modelzoo-services. I would like some guidance on the possible root cause and recommended troubleshooting steps.
Environment
Host OS: Windows 11
Board: STM32N6570-DK
Tools and versions:
- STM32CubeProgrammer: v2.20.0
- STEdgeAI: 2.2
- cubeIDE:1.19.0
Python: 3.9
Model and dataset
Model: st_yolo_x
Dataset: MS COCO
Completed steps:
Followed the official instructions to install and configure stm32ai-modelzoo-services on Windows
Successfully ran chain_tqeb (training / quantization / evaluation / benchmark) to completion
The quantization output folder contains:
Quantized model artifacts (e.g. .tflite / .onnx as produced by the pipeline)
A user_config file adapted from the official examples
Problem description
When running chain_qd or the deployment step for the STM32N6570-DK, the process fails at the “Board programming” stage.
The exception I receive is:
common.stm32ai_local.utils.STMAICToolsError: E400: Board programming failed!At the moment of failure, res_code = -1.
From the logs it looks like all model‑related steps (conversion, code generation, project creation) succeed, and only the board programming / flashing step fails.
Attachments
I attached a ZIP file that contains:
The quantized .tflite model generated from the st_yolo_x network trained on the MS COCO dataset.
This file reflects the exact model I am trying to deploy to the STM32N6570-DK when the E400: Board programming failed! res_code = -1 error occurs.
I will also test with different STM32CubeProgrammer versions on Windows to see whether the programmer version has any impact on this error and will update this thread with my findings.
