AI & ML


STM32Cube.AI upgraded to version 8

29 March 2023 AI & ML

STM32Cube.AI is a free tool for Edge AI developers that allows for the optimisation and deployment of trained neural network models from the most popular AI frameworks on any STM32 microcontroller. The tool is available via a graphical interface in the STM32CubeMX environment as well as in command line. This tool is now also available online in the STM32Cube.AI Developer Cloud.

The recently released version 8 introduces support for the ONNX Tensor-oriented file format (QDG). The following embedded AI frameworks have also been updated:

• TensorFlow 2.11.

• Keras.io 2.11.

• ONNX Runtime 1.12.1.

For deployment, either a trained neural network model can be loaded, or a model can be chosen from the STM model zoo, which is a collection of pre-trained ML models that are optimised to run on STM32 MCUs. The zoo is a resource available to anyone via GitHub, which allows developers to add AI capabilities to projects.




Share this article:
Share via emailShare via LinkedInPrint this page

Further reading:

World’s most powerful open LLM
AI & ML
With a staggering 180 billion parameters, and trained on 3,5 trillion tokens, Falcon 180B has soared to the top of the Hugging Face Leaderboard for pretrained LLMs.

Read more...
Bridging the gap between MCUs and MPUs
Future Electronics Editor's Choice AI & ML
The Renesas RA8 series microcontrollers feature Arm Helium technology, which boosts the performance of DSP functions and of AI and machine learning algorithms.

Read more...
Hardware architectural options for artificial intelligence systems
NuVision Electronics Editor's Choice AI & ML
With smart sensors creating data at an ever-increasing rate, it is becoming exponentially more difficult to consume and make sense of the data to extract relevant insight. This is providing the impetus behind the rapidly developing field of artificial intelligence.

Read more...
xG26 sets new standard in multiprotocol wireless device performance
Altron Arrow AI & ML
Silicon Labs has announced its new xG26 family of Wireless SoCs and MCUs, which consists of the multiprotocol MG26 SoC, the Bluetooth LE BG26 SoC, and the PG26 MCU.

Read more...
SolidRun unveils new SoM
Altron Arrow AI & ML
SolidRun and Hailo has unveiled a game-changer for engineers and AI product developers with the launch of their market-ready SoM, which packs the cutting-edge capabilities of the Hailo-15H SoC.

Read more...
Banana Pi with NPU
CST Electronics AI & ML
The latest Banana Pi SBC, the BPI-M7, is powered by Rockchip’s latest flagship RK3588 octa-core 64-bit processor, with a maximum frequency of 2,4 GHz.

Read more...
ESP32-P4 high-performance MCU
iCorp Technologies AI & ML
Powered by a dual-core RISC-V CPU running up to 400 MHz, ESP32-P4 also supports single-precision FPU and AI extensions, thus providing all the necessary computational resources.

Read more...
AI-native IoT platform launched
EBV Electrolink AI & ML
These highly-integrated Linux and Android SoCs from Synaptics are optimised for consumer, enterprise, and industrial applications and deliver an ‘out-of-the-box’ edge AI experience.

Read more...
Flash for AI
EBV Electrolink AI & ML
SCM offers a midway latency point between DRAM and SSDs, and when coupled with the introduction of CXL, low-latency flash, such as XL-FLASH, is well-positioned to deliver improvements in price, system performance, and power consumption to everything from servers to edge devices deploying the power of AI.

Read more...
Speeding up the rollout of renewable energy with AI
AI & ML
Understanding that AI, particularly within the renewables space, will not take away jobs, but rather create them, is key to leveraging the immense power of this technology to drive South Africa forward.

Read more...