STMicroelectronics has just announced the STM32Cube.AI Developer Cloud opening access to a suite of online AI development tools for the STM32 microcontrollers (MCUs) allowing developers to generate, optimize, and benchmark AI working on the company’s 32-bit Arm microcontrollers.
The company sus the STM32Cube.AI Developer Cloud is based on the existing STM32Cube.AI ecosystem of desktop tools with the added benefit of being able to remotely benchmark models on STM32 hardware through the cloud in order to save on workload and cost.
Some of the highlights of the online tools include:
- An online interface to generate optimized C-code for STM32 microcontrollers without requiring prior software installation.
- Access to the STM32 model zoo, a repository of trainable deep-learning models and demos. It currently features human motion sensing for activity recognition and tracking, computer vision for image classification or object detection, audio event detection for audio classification, and more. You’ll find those on GitHub along with getting started documentation.
- Access to an online benchmarking service for edge-AI Neural Networks on actual STM32 boards hosted in the cloud. The board farm is said to feature a broad range of STM32 boards used to remotely measure the actual performance of the optimized models.
You can run an AI workload in five steps:
- Register/login to the STM32Cube.AI Developer Cloud @ https://stm32ai-cs.st.com which is now freely available to any registered MyST user.
- Load your AI model which can be TFLite, Keras, ONNX, etc… or get one from the STM32 model zoo
- Set optimization options
- Benchmark your AI model on real STM32 MCU boards hosted in the cloud
- Generate and download C-optimized code for your own board.
STM32Cube.AI Developer Cloud has apparently been in private beta for a while as the email press release we’ve received includes quotes from various companies such as Zebra Technologies Corporation, Schneider Electric, Husqvarna Group AI Labs, Somfy, Lacroix and SIANA Systems who all moved from the desktop version of the tool to evaluate the cloud-based version.
The documentation and more details can be found on the product page.
Jean-Luc started CNX Software in 2010 as a part-time endeavor, before quitting his job as a software engineering manager, and starting to write daily news, and reviews full time later in 2011.