Company news
STM32Cube AI development tool adds depth quantization neural network support
Source:SPARKLAND ELECTRONICS CO.,LIMITED Time:2022-11-15
STM32Cube AI version 7.2.0, which is the first AI development tool launched by a microcontroller manufacturer to support ultra efficient deep quantitative neural networks.
STM32Cube. AI converts the pre trained neural network into C language code that can be run by STM32 microcontroller (MCU). It is an important tool to fully utilize the limited memory capacity and computing power of embedded products to develop cutting-edge AI solutions. Moving AI down from the cloud to edge devices can bring huge advantages to applications, including native privacy protection, deterministic real-time response, higher reliability and lower power consumption. Edge AI also helps optimize cloud computing usage.
Now, by supporting deep quantization input formats such as qKeras or Larq, developers can further reduce the amount of neural network code, memory usage and response delay. These advantages enable edge AI to release more possibilities, including economic applications and cost sensitive applications. Therefore, developers can create edge devices, such as self powered IoT endpoints with advanced functions and performance and longer battery life. From ultra-low power consumption Arm Cortex MCU ® From micro controllers to high-performance products using Cortex-M7, M33 and Cortex-A7 cores, STM32 series of STM32 of STF Semiconductor provides developers with many suitable hardware platforms.
STM32Cube. AI version 7.2.0 also added support for TensorFlow 2.9 model, improved kernel performance, and added scikit-learn machine learning algorithm and ONNX operator.
STM32Cube. AI converts the pre trained neural network into C language code that can be run by STM32 microcontroller (MCU). It is an important tool to fully utilize the limited memory capacity and computing power of embedded products to develop cutting-edge AI solutions. Moving AI down from the cloud to edge devices can bring huge advantages to applications, including native privacy protection, deterministic real-time response, higher reliability and lower power consumption. Edge AI also helps optimize cloud computing usage.
Now, by supporting deep quantization input formats such as qKeras or Larq, developers can further reduce the amount of neural network code, memory usage and response delay. These advantages enable edge AI to release more possibilities, including economic applications and cost sensitive applications. Therefore, developers can create edge devices, such as self powered IoT endpoints with advanced functions and performance and longer battery life. From ultra-low power consumption Arm Cortex MCU ® From micro controllers to high-performance products using Cortex-M7, M33 and Cortex-A7 cores, STM32 series of STM32 of STF Semiconductor provides developers with many suitable hardware platforms.
STM32Cube. AI version 7.2.0 also added support for TensorFlow 2.9 model, improved kernel performance, and added scikit-learn machine learning algorithm and ONNX operator.
86-755-23603356
SPARKLAND ELECTRONICS CO.,LIMITED
SPARKLAND ELECTRONICS CO.,LIMITED
Copyright:
Copyright:SPARKLAND ELECTRONICS CO.,LIMITED
Copyright:SPARKLAND ELECTRONICS CO.,LIMITED
Address:
Room 20E Bank of Communications, Century Place, 3018 Shennan Road Central, Futian District, Shenzhen
Room 20E Bank of Communications, Century Place, 3018 Shennan Road Central, Futian District, Shenzhen