Investment Rating - The industry is rated positively, with expectations of significant growth driven by AI demand and capital expenditures [28][29]. Core Insights - TSMC's AI guidance and CAPEX are above expectations, with projected revenue for Q1 2026 between $34.6 billion and $35.8 billion, and a gross margin of 63-65% [1]. - The company anticipates a nearly 30% year-over-year revenue growth for 2026, with a compound annual growth rate (CAGR) of 25% for revenue from 2024 to 2029, and 55-60% CAGR for AI processor revenue [1]. - The demand for AI is confirmed as genuine, leading to increased production capacity and a planned CAPEX of $52-56 billion for 2026 [1]. - Major North American cloud service providers (CSPs) are expected to increase their AI infrastructure investments, potentially reaching a total investment of $600 billion by 2026 [1]. - The report highlights strong demand for AI-related products, particularly in the PCB and semiconductor sectors, with companies actively expanding production [4][28]. Summary by Sections 1. Consumer Electronics - The report emphasizes the ongoing expansion of AI applications in consumer electronics, particularly in the Apple supply chain and smart glasses [5]. - AI mobile applications are expected to drive significant growth, with a focus on products like foldable phones and AI glasses [5]. 2. PCB - The PCB industry is maintaining high demand, particularly for copper-clad laminates, with expectations of price increases due to tight supply [6][28]. - The report indicates a robust outlook for AI-PCB companies, with strong orders and production capacity utilization [4][28]. 3. Semiconductor - The semiconductor sector is projected to benefit from increased demand for AI and cloud computing, with a positive outlook for memory chips and DRAM prices [22][24]. - The report notes that the semiconductor equipment and materials sectors are also expected to see growth due to geopolitical factors and domestic production initiatives [25][27]. 4. Key Companies - Notable companies mentioned include TSMC, NVIDIA, and Micron, which are expected to benefit from the AI-driven demand surge [28][29]. - Specific companies like North Huachuang and Zhongwei are highlighted for their advancements in semiconductor equipment and technology [30][31]. 5. Market Trends - The report indicates a strong upward trend in the semiconductor and PCB markets, driven by AI and cloud computing demands [28][37]. - The overall sentiment in the electronics sector is positive, with significant growth expected in the coming quarters [37].

Core Viewpoint - TSMC is increasing capital spending in response to strong demand for AI accelerators, indicating confidence in the AI market despite concerns about a potential bubble [1][3][10] Financial Performance - TSMC reported a 20.5% year-over-year revenue increase and a 35% jump in net income for Q4 2025, with expectations of nearly 30% revenue growth in 2026 and a 25% compound annual growth rate through 2029 [2][8] AI Demand and Capital Expenditure - The demand for AI accelerators is projected to grow by at least 50% annually through 2029, prompting TSMC to ramp up capital expenditures to between $52 billion and $56 billion in 2026, compared to around $40 billion in 2025 [3][8] - TSMC is also accelerating the start of production at its second fab in Arizona, with high-volume manufacturing expected to begin in the second half of 2027 [9] Market Confidence and Customer Insights - TSMC's CEO has engaged with customers to assess the sustainability of AI demand, concluding that "AI is real" and identifying it as an "AI megatrend" [6] - The financial health of hyperscale cloud providers, who are significant customers, has contributed to TSMC's confidence in future demand [6] Operational Efficiency - TSMC is utilizing AI to enhance productivity in its fabs, achieving gains of 1% to 2% at minimal cost [7] Industry Risks and Considerations - Despite the optimistic outlook, TSMC remains cautious about overexpansion, as the sustainability of AI demand and the financial viability of AI companies are uncertain [4][10][12] - The potential for unsustainable spending by AI companies raises questions about the long-term demand for AI computing capacity and TSMC's manufacturing capabilities [12][13]

Core Viewpoint - The concern regarding the AI bubble seems to be diminishing, which aligns with previous insights on the semiconductor landscape [1] Group 1: Analyst Background - The analyst, Uttam, specializes in growth-oriented investment analysis, focusing on technology sectors such as semiconductors, artificial intelligence, and cloud software [2] - Uttam has experience working in Silicon Valley with major technology firms like Apple and Google, leading teams [2] - He co-authors a newsletter, The Pragmatic Optimist, which is frequently cited by prominent publications like the Wall Street Journal and Forbes [2] Group 2: Analyst's Position - The analyst holds a beneficial long position in shares of AMD and TSM, indicating a vested interest in these companies [3]

[Table_Page] 投资策略周报|建筑装饰 证券研究报告 [Table_Title] 建筑行业周报 海外持续景气叠加国产替代提速，继续看多洁净室 [Table_ 相关研究： DocReport] [Table_Summary] 核心观点： | 建筑行业周报:核聚变大会即 | 2026-01-11 | | --- | --- | | 将召开、存储芯片短缺持 | | | 续，继续推荐利柏特，重视 | | | 核聚变、洁净室、绿氢工程 | | | 标的 | | [联系人： Table_Contacts]张子峻 021-38003620 zhangzijun@gf.com.cn 识别风险，发现价值 请务必阅读末页的免责声明 1/14 [Table_PicQuote] 相对市场表现 | [Table_Grade] 行业评级 | 买入 | | --- | --- | | 前次评级 | 买入 | | 报告日期 | 2026-01-18 | [分析师： Table_Author]耿鹏智 SAC 执证号：S0260524010001 021-38003620 gengpengzhi@gf.com.cn 分析师： 尉凯旋 ...

TSMC has for decades provided a "Silicon Shield†to its namesake island. Now it sees both business and geopolitical reasons for going big in the U.S. ...

半导体 半导体 2026 年 01 月 18 日 投资评级：看好（维持） 行业走势图 数据来源：聚源 -19% 0% 19% 38% 58% 77% 2025-01 2025-05 2025-09 半导体 沪深300 相关研究报告 《AI 算力自主可控的全景蓝图与投资 机遇—行业投资策略》-2026.1.8 《供应链安全事件催化，半导体材料/ 设备自主可控有望提速—行业点评报 告》-2026.1.8 《半导体释放涨价信号，晶圆厂、存 储、模拟有望进入价格上行期—行业 点评报告》-2025.12.25 先进封装龙头积极抢滩布局，产业进入"扩产+提价" 新阶段 ——行业点评报告 陈蓉芳（分析师） 祁海超（联系人） chenrongfang@kysec.cn 证书编号：S0790524120002 qihaichao@kysec.cn 证书编号：S0790125070022 台积电：2026 年资本开支超预期，先进封装持续规模化投入 2026 年 1 月 15 日，台积电召开 2025Q4 交流会，对 2026 年的资本开支指引为 520-560 亿美元（2025 年 409 亿，显著上修至多 36.9%），其中先 ...

Investment Rating - The report suggests a positive investment outlook for the AI hardware sector, highlighting strong demand for AI computing power and significant capital expenditure growth from TSMC in 2026 [5][19]. Core Insights - TSMC's Q4 2025 revenue reached $33.7 billion, exceeding Bloomberg's consensus by 3.3%, with a gross margin of 62.3% and an adjusted net profit of $16.29 billion, surpassing expectations by 10.4% [2][11]. - The HPC segment remains the core growth driver for TSMC, accounting for 55% of Q4 revenue and showing a 48% year-over-year growth, which is expected to continue into 2026 [2][19]. - The SELF DRIVE Act is poised to significantly increase the annual exemption limit for autonomous vehicles from 2,500 to 90,000 units per manufacturer, facilitating large-scale deployment of Level 4 autonomous vehicles [4][17]. - AMD and Intel's server CPU inventories are reportedly sold out, with a projected price increase of up to 15% due to high demand from hyperscale cloud providers [4][18]. Summary by Sections TSMC Performance - TSMC's Q4 2025 revenue was $33.7 billion, with a gross margin of 62.3% and an adjusted net profit of $16.29 billion, all exceeding market expectations [2][11]. - For the full year 2025, TSMC's revenue is projected at $122.56 billion, with a gross margin of 59.9% and an adjusted net profit growth of 552.7% [2][11]. - The company plans a capital expenditure of $52-56 billion for 2026, a significant increase from $40.9 billion in 2025, with expected revenue growth of nearly 30% year-over-year [12][19]. AI Hardware Demand - The report emphasizes strong and certain demand for AI computing power, with TSMC's HPC business expected to account for 58% of total revenue in 2025, growing at 48% year-over-year [5][19]. - The introduction of Anthropic's Cowork feature marks a significant advancement in AI applications, allowing for more autonomous task management and collaboration [3][14]. Autonomous Vehicle Regulations - The SELF DRIVE Act aims to enhance the regulatory framework for autonomous vehicles, proposing to raise the exemption limit for manufacturers significantly, which could accelerate the deployment of autonomous vehicle technology [4][17]. - The bipartisan support for the SELF DRIVE Act indicates a strong political will to advance autonomous vehicle technology in the U.S. [4][17]. Server CPU Market - The server CPU market is experiencing a supply-demand imbalance, with AMD and Intel's inventories reportedly sold out, leading to anticipated price increases of up to 15% [4][18]. - The demand surge is primarily driven by hyperscale cloud providers upgrading their server architectures [4][18].

证券研究报告 | 行业周报 gszqdatemark 2026 01 18 年 月 日 电子 周观点：从台积电业绩看全球 AI 需求爆发 台积电：FY25Q4 业绩超指引，AI 需求延续强劲。台积电 25Q4 业绩表 现超预期，营收与利润率同步上行，主要受先进制程需求强劲带动。其中 3nm/5nm制程占比持续提升，7nm及更先进制程贡献已接近八成，FY26Q1 指引延续强势。从应用端来看，HPC 仍为最大增长引擎。管理层预计 Foundry2.0增长稳健并将持续跑赢行业，公司全年收入有望维持接近30% （美元口径）的高增长，同时强调 AI 需求" 真实健康"、可见度提升。产 能布局方面，公司加速美国、日本及中国台湾等地扩产规划，以匹配 AI、 HPC 等长期趋势，并通过节点优化与先进封装投入提升供给弹性。技术路 线方面，N2 已进入量产并预计 2026 年快速爬坡，N2P 与 A16 将于 2026 年下半年陆续量产，N2 family 有望成为新一代长生命周期平台。同时，公 司上调 AI 加速器业务的收入增长预测，预计 2024-2029 年五年期间的 CAGR 将达到 mid-to-high 50%，可见 ...

2026 年 01 月 18 日 行业研究●证券研究报告 http://www.huajinsc.cn/ 1 / 10 请务必阅读正文之后的免责条款部分 电子 行业周报 AI &半导体：台积电大幅提升 2026 年资本开支 投资要点 投资评级 领先大市(维持) 首选股票 评级 一年行业表现 资料来源：聚源 升幅% 1M 3M 12M 相对收益 12.16 10.7 40.31 绝对收益 17.37 13.15 64.83 分析师 熊军 SAC 执业证书编号：S0910525050001 xiongjun@huajinsc.cn 相关报告 电子：AI &半导体：Marvell 拟战略收购 Celestial AI-华金证券-电子-周报 2025.12.7 电子：AI &半导体：阿里巴巴 AI 产品收入持续 高增长-华金证券-电子-周报 2025.11.29 电子：AI &半导体：NAND 市场供不应求趋势 有望持续-华金证券-电子-周报 2025.11.16 台积电发布 2025 年第四季度业绩。公司公布 Q4 营收 337.3 亿美元，同比增长 25.5%，环比增长 1.9%。毛利率 62.3%，同比提升 3 ...

1）台积电资本支出及业绩超预期：A）台积电公布 2025 年 Q4 财务报告，公司当季营收为 337.3 亿美元，同比+25.5%， 环比+1.9%。全面上修业绩增速与毛利率指引，将 2024 至 2029 年 AI 营收 CAGR 从原先的"40%中段"提升至"50%中 高段"。B）2026 年资本支出规模进一步扩大至 520 亿至 560 亿美元。公司管理层强调，约七成至八成资本将投向先 进制程技术，其余用于特殊制程、先进封装等领域，以应对持续旺盛的 AI 芯片需求。全球 AI 算力需求持续旺盛。2） 国产芯片、国产大模型、国产应用携手共进，加速国产链向上：A）首个完全依托国产芯片全程训练的 SOTA 多模态模 型 GLM-Image，开源 24 小时内登顶全球知名 AI 开源社区 Hugging Face Trending，标志国产 AI 模型的端到端自主研 发能力在国际竞争中取得突破。B）DeepSeek 发布新论文，提出 Engram 机制，在等参数、等算力条件下显著提升模型 在知识调用、推理、代码、数学等任务上的表现。同时，DeepSeek 开源相关记忆模块 Engram。我们认为此机制意味 着在 ...