The Switzerland-based European semiconductor corporation STMicroelectronics (ST) announced a 'developer-first' strategy in line with the spread of artificial intelligence (AI) and unveiled new low-power, high-performance microcontrollers (MCUs). Acting as the brain of electronic devices, MCUs perform various functions, such as sensor control, data processing, and communication. Recently, as the demand for low-power, high-efficiency chips has increased during the inference stage of AI development, the demand for high-performance MCUs has also been on the rise.
Choi Kyunghwa, Marketing Director of ST Korea, noted at a press conference held on the 20th at the Novotel Ambassador in Gangnam, Seoul, that 'the number of global developers using ST's universal MCU product line STM32's software development tools reached 1.3 million last year,' adding that 'based on a high user satisfaction rate of over 80 points, the STM32 has maintained the number one position in the universal MCU market for three consecutive years.' Since its launch in 2007, more than 13 billion STM32 MCUs have been sold, establishing itself as the standard MCU across various industries.
To respond to the growing demand for edge AI (AI that processes data on devices used by individuals rather than large computers), ST has introduced high-performance MCUs designed to rapidly process AI computations even in small devices. The 'STM32N6' combines a low-power MCU with ST's self-developed neural processing unit (NPU) to enhance AI inference performance. It processes more than 600 billion multiplication operations per second, improving AI computation capabilities by over 600 times compared to existing products. It can implement edge AI at low power in various devices such as drones, smart homes, robots, and automobiles.
Moon Hyunsoo, Director of Edge AI Solutions at ST Korea, explained that 'while high-performance GPUs are needed in the learning stage, inference can be sufficiently implemented even on small embedded platforms (computer chips or systems operating with limited resources),' adding that 'the STM32N6 is an MCU optimized for machine learning algorithms through a self-developed neural art accelerator.'
Alongside this, ST launched new products that strengthen wireless consolidation and security features. The low-power short-range wireless MCU 'STM32WBA6' supports multiple wireless communication methods, including low-power Bluetooth, Thread, Zigbee, and Matter. Equipped with dual-bank flash memory, wireless firmware updates can be conducted without application interruptions. It has obtained global security certifications such as the U.S. Cyber Trust Mark, enabling compliance with cybersecurity regulations in various countries.
The ultra-low-power MCU 'STM32U3,' characterized by high energy efficiency, is specialized for products such as industrial medical devices and smartwatches, where battery life and security are crucial. It applies technologies that precisely adjust the minimum voltage required for stable device operation, minimizing power consumption during operation. With a built-in security accelerator, it can safely protect sensitive data.
Director Choi stated, 'It is expected that over 60% of MCU products will transition to products based on fine process technology by 2030, and ST is also expanding its product lineup that utilizes fine processes.' He added, 'As the growth rate of the STM32 product line is expected to be 1.5 times faster than the market average, we will strengthen our market competitiveness based on a developer-centered strategy and a strong ecosystem.'