Guoyuan Securities·2026-02-02 08:11Investment Rating - The report maintains a "Buy" recommendation, emphasizing an AI-driven upward economic cycle and the search for structural investment opportunities in the electronics industry [1]. Core Insights - The overall upward cycle in 2026 is expected to continue, with AI infrastructure yet to show significant demand-driving effects [4]. - The A-share technology companies are currently in an upward economic cycle, having entered a recovery phase in January 2024 and a prosperous phase in August 2024 [6][11]. - The semiconductor sector is expected to outperform the consumer electronics sector, driven by AI [11]. Industry Performance - The semiconductor and storage sectors are experiencing a cyclical recovery, with AI's focus shifting from infrastructure to connectivity and operational capacity [12]. - The report highlights significant stock price increases for various companies, such as Micron (up 240.1%), Lam Research (up 139.2%), and Intel (up 84.0%) from January 1, 2025, to December 31, 2025 [12]. - A-share technology companies like Industrial Fulian and Huahong Semiconductor have also shown substantial growth, with increases of 194% and 132%, respectively [13]. Market Dynamics - The global semiconductor market has emerged from the previous downturn, with a positive growth cycle expected to persist into 2026 [19][22]. - The report notes that the main growth drivers are AI and rising prices of storage chips, with the industry entering a comprehensive upward trend from August 2024 [26]. - The demand for AI-related products and services is anticipated to increase, particularly in the context of data interconnectivity and GPU localization [43][80]. Downstream Demand - The report indicates that while AI infrastructure remains strong, it has not yet led to a clear increase in downstream demand [39]. - The overall inventory levels in the industry are low, and the demand outlook is still weak, suggesting a cautious approach to investment [39]. AI and Semiconductor Trends - The report identifies two major directions for AI: high-speed PCB and upstream domestic AI GPUs [41]. - The demand for domestic AI GPUs is expected to grow significantly in 2026, although there remains a substantial capacity gap [80]. - The report emphasizes the importance of GPU acceleration and the anticipated explosive growth in GDDR7 demand [64].