海外科技行业2026年第4期：AI从算力走向能源，基础设施逻辑持续强化

Investment Rating - The report maintains an "Overweight" rating for the industry, recommending investment in AI computing, cloud vendors, AI applications, and AI social sectors [4]. Core Insights - The report highlights that the AI industry is transitioning from being "compute-constrained" to "energy-constrained," with commercialization paths for robotics, autonomous driving, and space computing infrastructure accelerating [5][6]. - Elon Musk's insights during the World Economic Forum emphasize the importance of energy supply over chip availability for AI development, noting China's advantages in power infrastructure and photovoltaic industries [5]. - Jensen Huang, CEO of NVIDIA, refutes concerns about an AI bubble, asserting that the ongoing investment cycle in infrastructure will require trillions of dollars over the coming years, which is essential for supporting the entire AI ecosystem [5][6]. Summary by Sections Investment Recommendations - The report suggests maintaining an "Overweight" rating for the industry and recommends specific sectors: AI computing, cloud vendors, AI applications, and AI social platforms [4][24]. Company Performance - Netflix's Q4 2025 revenue reached $12.05 billion, a year-on-year increase of 17.6%, with operating profit and net profit growing by 30.1% and 29.4%, respectively [5]. - TSMC reported Q4 2025 revenue of $33.7 billion, with a gross margin of 62.3%, reflecting strong demand for 3nm technology [7]. Market Trends - OpenAI is introducing advertising in its free and entry-level subscription versions of ChatGPT, aiming to monetize its user base amid significant operational losses [8]. - Major DRAM manufacturers, including Samsung and SK Hynix, are increasing production, but supply will still fall short of market demand, indicating a continued storage supercycle [9]. Industry News - Suir Technology's IPO on the STAR Market aims to raise 6 billion yuan for AI chip development, reflecting ongoing commercialization efforts in the AI sector [22]. - Micron Technology has reported a significant shortage of storage chips, exacerbated by rising demand for high-end semiconductors in AI infrastructure [22].