MCU市场，变天了

Core Insights - The rise of artificial intelligence (AI) is transforming the demand for microcontrollers (MCUs), driven by the need for AI-generated code and inference, as well as cybersecurity standards [1] - The transition to advanced 22nm process technology is enabling new memory technologies like MRAM to deliver higher performance at lower costs, facilitating the entry of new MCU architectures into the market [1] Memory Safety - SCI Semiconductor has introduced its ICENI security 32-bit microcontroller, the first commercial device using the CHERI (Capability Hardware Enhanced RISC Instructions) secure memory architecture [2] - The ICENI device combines the RISC-V RV32E instruction set with CHERI hardware architecture, enhancing the security of legacy and AI-generated code with minimal recompilation effort [2] - This memory safety approach can eliminate 70% of critical vulnerabilities and improve system resilience with less than 2% impact on code performance [2] Production and Independence - ICENI microcontrollers are manufactured using GlobalFoundries' low-power 22FDX SOI process, ensuring European autonomy and enabling production in the U.S. [3] - Nordic Semiconductor is facing similar challenges in maintaining independence as a microcontroller supplier while expanding its software ecosystem through acquisitions [9] New Product Launches - Silicon Labs is preparing to launch its Series 3 MCU, which features a 22nm platform allowing customer software to run in parallel with wireless stacks [5] - The MSPM0G5187 and AM13Ex MCUs from Texas Instruments integrate the TinyEngine NPU, significantly reducing latency and energy consumption for edge processing [12] - STMicroelectronics is lowering the price of its STM32C5 microcontroller to $0.64, targeting applications in smart thermostats, electronic locks, and wearable devices [16] AI Integration - Ambient Scientific's GPX-10 AI microcontroller is being utilized in a wearable device designed for women's safety, featuring continuous AI functionality and a long battery life [14][15] - The integration of AI accelerators in microcontrollers is expected to enhance the intelligence of consumer electronics and improve efficiency in industrial applications [12]