3 6 Ke·2025-04-30 01:11Group 1: Acquisition of Deeplite by STMicroelectronics - STMicroelectronics (ST) has acquired Canadian AI startup Deeplite, which specializes in edge AI technology, particularly in model optimization, quantization, and compression [1][2] - Deeplite's technology enables AI models to run faster, smaller, and more energy-efficiently on edge devices, addressing significant challenges in deploying deep learning models commercially [2][4] - The acquisition is expected to enhance ST's STM32N6 high-performance microcontroller adoption, leveraging Deeplite's automated software engine for optimizing deep neural networks [2][5] Group 2: Edge Impulse Acquisition by Qualcomm - Qualcomm announced its acquisition of Edge Impulse, an edge AI development platform, to expand its AI capabilities for IoT products [6][7] - The acquisition is anticipated to accelerate support for Qualcomm's Dragonwing processors while maintaining Edge Impulse's brand and platform accessibility for various hardware partners [6][7] - Edge Impulse's platform is widely adopted for adding AI functionalities to embedded systems, with significant applications in health wearables and industrial organizations [7][8] Group 3: NXP's Acquisition of Kinara - NXP has reached an agreement to acquire Kinara, a leader in high-performance and energy-efficient discrete neural processing units (NPU), for $307 million [10][11] - Kinara's NPUs are designed for a wide range of edge AI applications, supporting multimodal generative AI models and ensuring adaptability for future AI algorithm developments [11][12] - The acquisition is expected to be completed by mid-2025, pending regulatory approvals [10] Group 4: Trends in Edge AI - The trend towards edge AI is growing, with predictions indicating that by 2025, 75% of data will be processed at the edge, highlighting the market potential for edge AI microcontrollers [14][15] - Major MCU manufacturers are actively acquiring startups in the edge AI space, indicating a rapid increase in demand for edge AI computing [14][15] - The competitive landscape among MCU manufacturers is expected to intensify as they adapt to the growing need for embedded AI/ML solutions [15]