20250916-开源证券-电子行业点评报告_AI大模型厂商加速导入硬件入口_端侧AI产业链投资机遇可期_3页_503kb

报告摘要

AI End-Device Industry Analysis Summary

Overview of AI Model Evolution

AI models are advancing with a focus on edge computing and lightweighting. Global models like GPT series and Gemini iterate quickly through algorithm optimization and data accumulation, enhancing capabilities such as code generation and inference efficiency. This enables better performance for edge deployment, reducing latency and costs by balancing cloud and terminal computing.

Importance of Hardware Integration

The integration of AI into hardware devices is becoming critical. Major AI model providers, such as Google, are partnering with device manufacturers in hardware entry points, enabling real-time interactions and various applications. Examples include Google's Gemini in Samsung devices, Alibaba's collaboration with Honor for services like navigation and shopping, and Apple's upcoming high-level AI features. This trend is expected to drive AI market growth in terminals by 2025.

AI Terminal Hardware Iterations

Terminal devices are evolving to support AI with enhancements in processing power, energy efficiency, and form factors. Hardware includes SoC-integrated NPUs for local computation and advanced storage for faster data access. Energy efficiency improvements involve better cooling and batteries, while new device forms like AI glasses and earphones expand application possibilities.

Investment Opportunities

Key investment targets include: (1) Brand devices, e.g.,传音控股、小米集团-W、立讯精密、歌尔股份、旗帜科技、华勤技术; (2) Components suppliers, e.g.,领益智造、蓝思科技、鹏鼎控股、瑞声科技、舜宇光学科技、电连技术、水晶光电、蓝特光学、珠海冠宇; (3) Chips and storage providers, e.g.,瑞芯微、恒玄科技、乐鑫科技、中科蓝讯、炬芯科技、汇顶科技、兆易创新、艾为电子、格科微、思特威、唯捷创芯、美芯晟.