Welcome Here  Shenzhen Mingjiada Electronics Co., Ltd.

Recently, NXP announced a new i.MX RT700 crossover MCU family to support smart AI for edge-side devices such as wearables, consumer medical devices, smart home devices and HMI platforms. According to officials, the new MCUs are designed to deliver significant power savings and can provide up to 172x AI acceleration at the edge end.

MCU integrated NPU supports AI functionality

NXP launched this i.MX RT700 internal integration of NXP self-developed eIQ Neutron NPU, a scalable hardware accelerator architecture designed for AI and ML applications. the introduction of the NPU dramatically improves the performance and energy efficiency of the device's AI workloads, making it possible to deploy AI models locally.

MCU integration of NPUs is an important trend in current technology development. With the popularity of AI technology and the increase in demand for edge computing, traditional MCUs have struggled to meet the demand for efficiently processing complex AI tasks. Therefore, integrating NPUs into MCUs has become one of the key technologies to improve the level of device intelligence and response speed.

Designed for neural network computing, NPUs can significantly increase the computing speed and efficiency of MCUs when processing AI algorithms. This enables MCUs to perform more complex AI tasks such as image recognition, speech recognition, natural language processing, and so on.

Compared to traditional CPUs or GPUs, NPUs have lower power consumption when processing neural network operations. This is especially important for devices that need to run for long periods of time, such as wearables and smart home devices.

In the smart home field, MCUs with integrated NPUs are able to process sensor data in real time to achieve intelligent linkage and scene control of smart home devices. For example, through image recognition technology, smart door locks can identify the owner's facial features to achieve keyless entry; intelligent lighting systems can automatically adjust the brightness and colour temperature according to the indoor light and personnel activities.

In the field of industrial automation, MCUs with integrated NPUs can perform more complex machine vision and machine learning tasks, such as product quality inspection and production process monitoring. This helps to increase productivity, reduce labour costs and improve product quality.

In the field of medical devices, MCUs with integrated NPUs can process medical image data in real time to assist doctors in disease diagnosis and treatment decisions. For example, in intelligent diagnostic systems, NPUs can quickly analyse patient image data and provide preliminary diagnostic recommendations.

In the field of automotive electronics, MCUs with integrated NPUs can implement advanced driver assistance systems (ADAS) and autonomous driving functions. For example, through image recognition and sensor fusion technology, cars can perceive the surrounding environment in real time and make driving decisions accordingly.

At present, a number of MCU manufacturers have launched products with integrated NPUs. In addition to NXP, STMicroelectronics and Infineon have launched MCUs with integrated NPUs, such as STMicroelectronics' STM32N6 series, which is the first of its MCUs equipped with an NPU, which enables the series of MCUs to execute complex AI algorithms at the edge end, greatly improving processing power and efficiency.

According to Remi El-Ouazzane, president of ST's Microcontrollers, Digital ICs and RF Products Division, the STM32N6 delivers the same AI performance as a quad-core processor equipped with an AI accelerator, but at one-tenth the cost and one-twelfth the power consumption. This is largely due to its efficient architectural design and optimised power management.

The STM32N6 series, due to its powerful AI processing capabilities and low-power characteristics, is well suited for edge computing scenarios with stringent requirements on power consumption and performance, e.g., in smart cities and industrial IoT, the STM32N6 can be used as the core processing unit of smart sensors to realise real-time data acquisition, processing and transmission.

Using its built-in NPU, STM32N6 can perform machine vision tasks such as image recognition, object detection, etc., providing technical support for intelligent security, autonomous driving and other fields. In embedded systems, the STM32N6 can significantly improve the intelligence of devices and equip them with more powerful data processing and decision-making capabilities.

Then there is Infineon, which released its new new PSOC Edge E81, E83 and E84 microcontrollers in April this year, equipped with the high-performance Arm Cortex-M55 core, supporting the Arm Helium DSP and paired with the neural network processor of the Arm Ethos-U55, and the Cortex-M33 core paired with Infineon's ultra-low-power NNLite. NNLite. NNLite is a proprietary hardware accelerator for accelerating neural networks.

Optimised for machine learning applications running on interconnected systems at the edge of the network, these fully integrated system-on-chips enable designers to bring advanced AI capabilities to IoT and consumer-grade applications.

Target applications for PSOC Edge MCUs include human-machine interfaces (HMIs) in appliances and industrial equipment, smart home and security systems, robotics and wearables. These devices require high performance, low power consumption and robust security to support complex machine learning applications and user interaction experiences.

Company Home: http://www.hkmjd.com/