Group 1 - The core point of the article is that Taiwan Semiconductor Manufacturing Company (TSMC) reported a revenue of approximately NT$330.98 billion for September 2025, which represents a 1.4% decrease from the previous month but a 31.4% increase compared to the same month last year [1][2]. Group 2 - In September 2025, TSMC's revenue was NT$330.98 billion, down from NT$335.77 billion in August 2025 [2]. - Year-over-year, the revenue increased from NT$251.87 billion in September 2024, marking a growth of 31.4% [2]. - For the first nine months of 2025, TSMC's revenue reached NT$2,762.96 billion, which is a 36.4% increase compared to NT$2,025.85 billion in the same period of 2024 [2].

TSMC , the world's largest contract chipmaker, reported on Thursday third-quarter revenue of T$989.92 billion ($32.48 billion), beating market forecasts, and up 30% on the year ago period on surging i... ...

市场有风险，投资需谨慎。本文不构成个人投资建议，也未考虑到个别用户特殊的投资目标、财务状况或需要。用户应考虑本文中的任何 意见、观点或结论是否符合其特定状况。据此投资，责任自负。 台积电：9月销售额3,309.8亿元台币，同比增长31.4％。 风险提示及免责条款 ...

格隆汇10月9日｜巴克莱：将台积电(TSM.US)目标价从325美元上调至330美元。 ...

Core Insights - The article discusses the recent performance and developments in the tech sector, particularly focusing on AMD and its integration with OpenAI, as well as the advancements in AI models like Sora 2 [1][3][4]. Group 1: AMD and AI Integration - AMD has been included in a closed-loop AI ecosystem, which is seen as a positive development despite uncertainties regarding TSMC's production capabilities for 3nm and 2nm chips [1][3]. - The article highlights that traditional cloud companies may not participate in this closed-loop due to their conservative management styles and focus on stable returns [3]. Group 2: Sora 2 Model - The Sora 2 model, set to launch in February 2024, is compared to a significant advancement in video generation akin to GPT-3.5, capable of complex tasks such as simulating Olympic gymnastics movements [3]. - OpenAI's Sora 2 is noted for its improved controllability and ability to follow intricate instructions across multiple scenes while maintaining continuity in the generated content [3]. Group 3: Market Performance - The Sora app achieved the highest download volume during the National Day holiday, indicating strong market interest [4]. - The Hang Seng Tech Index ETF (513180.SH) has shown a year-to-date increase of 43%, with a notable rise of 34.7% since early October [9][13]. - The overall valuation of the Hang Seng Tech Index is significantly lower at 24.9 times earnings compared to the 204 times earnings for the STAR Market, suggesting a potential for catch-up in performance [13].

Core Insights - TSMC's 2nm wafer foundry price increase is significantly lower than market expectations, with a rise of 10% to 20% compared to the 3nm process, setting the price at approximately $30,000 per wafer, which is much lower than the previously rumored 50% increase, indicating a more stable pricing system for advanced processes [1][2] Pricing Strategy - TSMC's pricing strategy for the 2nm process features two main characteristics: the price increase is kept within a reasonable range, with a focus on a 15% to 20% rise compared to the average price of the 3nm process, alleviating cost pressure for clients adopting cutting-edge processes [1] - The company has established a two-way adjustment mechanism to balance costs and pricing, encouraging equipment and material suppliers to provide a 10% to 20% cost reduction, while also reasonably passing some cost pressure to the manufacturing side to ensure sustainable R&D investment and capacity building [1] Future Pricing Adjustments - TSMC plans to implement comprehensive price adjustments for its 3nm, 4nm, 5nm, and 7nm advanced processes next year, with increases expected to remain in the single-digit percentage range, with specific adjustments being flexibly determined based on client cooperation scale, enhancing adaptability to different client needs [1] Capacity Utilization - TSMC's advanced process capacity is currently fully utilized, reflecting strong market demand for high-end chip manufacturing, while mature process production lines are not fully utilized, allowing for future adjustments in industry demand structure [2] - The overall pricing system for the 2nm process, which includes multiple technology nodes like N3P and N3E, has matured, with major clients already signing orders for 2025, laying a solid foundation for the mass production of the 2nm process [2] Production Timeline - The mass production of TSMC's 2nm process is expected to break through in the second half of 2025, with the initial capacity focusing on serving high-performance computing and artificial intelligence chip clients [2]

Summary of Key Points from the Conference Call Industry and Company Overview - **Industry**: AI Supply Chain, specifically focusing on semiconductor manufacturing and AI ASICs - **Key Companies**: TSMC (Taiwan Semiconductor Manufacturing Company), NVIDIA, Meta, Google, AWS, Alchip, MediaTek, GUC (Global Unichip Corp) Core Insights and Arguments 1. **CoWoS Capacity and Demand**: - TSMC's 2026 CoWoS capacity plan is currently under-supplying based on anticipated demand from key customers, particularly NVIDIA [2][10] - NVIDIA's RTX Pro 6000 forecast remains strong, indicating robust inference demand in China [1][2] - TSMC's current offering of 590k CoWoS-L wafers in 2026 is projected to be 20% below NVIDIA's demand [2] 2. **NVIDIA's AI ASIC Developments**: - The Rubin CPX GPU is expected to adopt TSMC's CoWoS-S packaging, with a performance of 30 PFLOPS and 128GB of GDDR7 memory [3][39] - NVIDIA's AI capacity costs are estimated at $50-60 billion per GW, with $35-40 billion allocated to NVIDIA, suggesting potential revenues of $350-400 billion starting in 2H26 [2][11] 3. **AI ASIC Volume Forecast**: - AI ASIC volumes could reach approximately 5.7 million units in 2026 and 8 million units in 2027, with Google and AWS being the primary players [47] - MediaTek's chances of winning the Meta MTIA project are diminishing due to strong competition from Broadcom and Marvell [57][58] 4. **Market Dynamics and Competitor Analysis**: - GUC is expected to benefit from increased demand for Google Axion CPUs, potentially contributing around $250 million to its revenue in 2026 [59] - Alchip is projected to generate approximately $1.8 billion in revenue from Trainium3 in 2026, despite a back-loaded shipment schedule [56] 5. **Capacity Expansion and Future Projections**: - TSMC's clean room space is ready for expansion to 110-120k if demand is confirmed, with a revised CoWoS capacity assumption of 100kwpm by 2026 [2][12] - The overall demand for CoWoS is expected to grow significantly, with NVIDIA's demand forecasted to increase by 61% in 2026 [20] Additional Important Insights - **NVIDIA's Strategic Positioning**: - NVIDIA's Rubin architecture is designed to optimize inference performance, potentially delivering a 30x to 50x return on investment [39] - The demand for NVIDIA's Blackwell GPU, including RTX Pro 6000, is strong among Chinese customers, with expected sales of 1.5 million to 2 million units in 2H25 [14] - **Competitive Landscape**: - SMIC is aggressively expanding its 7nm node capacity to meet domestic GPU and AI ASIC demand in China [14] - The competition in the AI ASIC market is intensifying, with various players vying for market share and technological advancements [57][58] This summary encapsulates the critical insights and projections regarding the AI supply chain, particularly focusing on TSMC and NVIDIA's roles within the semiconductor industry.

Core Viewpoint - TSMC's advancement in wafer fabrication technology is leading to increased costs, impacting clients like Apple, Qualcomm, and MediaTek, although the premium for the new 2nm process may be less severe than initially expected, ranging from 10% to 20% compared to the 3nm process [1][2]. Pricing Dynamics - The anticipated price for TSMC's 2nm wafers is projected to be around $30,000 each, with the potential for a 50% price increase next year [5]. - TSMC's current 3nm process is expected to see price hikes, with the second-generation N3E reaching approximately $25,000 and the third-generation N3P around $27,000 [2][5]. Client Adaptation - Major clients like Qualcomm and MediaTek are adapting to the price increases, with Qualcomm planning to transition to the 2nm process for its Snapdragon 8 Elite Gen 6 SoC by 2026, and MediaTek already having successfully taped out its first 2nm SoC [3][5]. - Despite the cost pressures, clients are still willing to invest in TSMC's latest technology, indicating a strong demand for advanced semiconductor processes [3]. Market Demand and Competition - TSMC is reportedly experiencing high demand for its 2nm process, with 15 major companies preparing to adopt it, including AMD and MediaTek, and expectations that Apple will also become a client [5][6]. - The semiconductor industry is facing inflationary pressures, with rising prices for memory and storage components contributing to overall cost increases [5]. Production Capacity - TSMC is expanding its production capabilities by constructing multiple 2nm fabs in Taiwan and a third fab in Arizona, aiming to meet the growing market demand [6].

与此同时，根据SEMI周三发布的报告，得益于人工智能芯片的蓬勃发展，预计2026年至2028年，全 球用于生产12英寸（300毫米）半导体晶圆的芯片工厂设备的支出将达到3740亿美元，这些设备支持 成熟和先进的芯片制造。 公众号记得加星标⭐️，第一时间看推送不会错过。 来源 ：本文编译自日经 。 国际半导体产业协会 (SEMI) 预测显示，随着人工智能需求的激增以及华盛顿推动本地化生产，美国 半导体投资将从2027年起超过主要芯片经济体中国大陆、中国台湾和韩国。 2027年至2030年，美国的芯片投资（包括设备和设施建设支出）将大幅增长，部分原因是政策驱动 的对先进逻辑芯片和存储芯片的投资。 根据行业机构SEMI的预测，今年和明年美国芯片投资将达到约210亿美元，到2027年和2028年将分 别增至330亿美元和430亿美元。 SEMI市场情报高级总监Clark Tseng表示，预计2027年至2030年，中国的芯片总投资将达到约1580 亿美元，这一增长幅度在世界其他地区是无法比拟的。 Clark在亚利桑那州凤凰城举行的SEMICON West贸易展期间对《日经亚洲》表示："根据我们目前看 到的已确认的半导 ...

TSMC x Nvidia : Breaking the Thermal Wall: How Advanced Cooling Is Powering the Future of Computing 台积电 x Nvidia ：突破热壁：先进冷却技术如何驱 动未来计算 On the logic process side, TSMC's evolution from N3 → N2 → A16 involves more than geometric scaling—it represents a transistor architecture transition: from FinFET, to GAA (Gate-All-Around), and finally to the Super Power Rail architecture in the A16 generation. This series of changes is aimed at achieving the optimal balance of PPA (Power, Performance, and Area), laying the ...