证券研究报告 | 行业点评报告 2025 年 08 月 24 日 DS 再燃自主可控热情，关注国产 AI 算力芯片产业链 国产算力芯片链跟踪报告 TMT 及中小盘/电子+通信+计算机 2025 年以来北美 CSP Capex 进展趋势乐观，中美博弈大背景下 H20 等芯片销 售情况持续反复，国内算力芯片公司在供给受限和自主化需求背景下重要性持 续提升，本土 GPU 公司 2025 年销售同比持续增长且长期趋势乐观，半导体属 AI 算力和自主可控双重属性叠加领域，长期持续推荐 GPU/ASIC/Switch 和配 套芯片、先进制造/存储、上游设备/材料/EDA、PCB/CCL 等算力产业链。 推荐（维持） 行业规模 | | | 占比% | | --- | --- | --- | | 股票家数（只） | 506 | 9.9 | | 总市值（十亿元） | 11216.1 | 11.2 | | 流通市值（十亿元） | 9833.5 | 10.8 | 行业指数 % 1m 6m 12m 绝对表现 18.7 19.4 91.6 相对表现 13.7 11.6 62.5 资料来源：公司数据、招商证券 -20 0 20 40 6 ...

前言 财联社倾力打造王牌栏目《风口研报》，替您"八一八"市场含金量超高的研报、调研信息。以机构视 角，追踪研报和调研纪要细节里的"超预期"、"拐点"、"事件催化"和"价值洼地"。 AI服务器强劲需求拉动HBM指数级增长，到2030年全球HBM市场有望达980亿美元，24-30年复合增速 达33%，明年ASIC同比增50%成新需求增量，分析师看好提前布局产业链的公司。 ...

Semiconductor Industry Research Summary Industry Overview - The semiconductor industry is driven by Moore's Law, which states that the number of components on integrated circuits doubles approximately every 18 to 24 months, leading to reduced costs and expanded application scenarios, including IoT and brain-machine interfaces [1][6] - The global semiconductor market is expected to exceed $1 trillion by 2030, with integrated circuits being the main driver, accounting for 80% of the market, and digital chips making up 80% of integrated circuits [1][15] Key Points on Semiconductor Chips - Semiconductor chips are categorized into five functional types: information acquisition, transmission, processing, storage, and output [1][7] - Integrated circuits represent 80% of the semiconductor industry's value, with digital chips resembling the human brain, responsible for logic and information storage [1][8] - Digital chips include various types such as CPU, MCU, FPGA, GPU, DRAM/Flash, and ASIC/SoC, while analog chips manage signal chains and power distribution [1][10][12] Market Dynamics - The semiconductor market is primarily driven by consumer electronics, which account for 60%-70% of downstream applications, with mobile phones representing about 30% [1][16] - AI development is rapidly changing the market landscape, with NVIDIA's data center revenue nearing 25% of the semiconductor market, and AI-related semiconductors approaching 30% [1][16] Manufacturing and Design Processes - Semiconductor manufacturing involves design, fabrication, and testing, with critical processes including photolithography, etching, deposition, and ion implantation [1][4][17] - The semiconductor industry operates in a triangular structure, with product layers at the bottom, manufacturing layers vertically, and equipment and materials on the sides [1][20] Financial Aspects and Valuation - Chip design companies generate revenue based on sales volume multiplied by unit price, while wafer manufacturers rely on capacity, utilization rates, and pricing [1][22][23] - Valuation methods differ across semiconductor sectors, with design companies typically evaluated on PE ratios based on growth expectations, while wafer and testing companies are often assessed using PB ratios [1][26] Innovation and Growth Opportunities - Innovation cycles are crucial in the semiconductor industry, as they drive value growth across various applications, particularly in AI [1][28] - Identifying high-quality semiconductor companies involves analyzing end-user growth rates and changes in chip value, particularly in emerging sectors like electric vehicles and photovoltaics [1][29] Investment Considerations - Key investment points in the semiconductor industry include innovation, new directions, and understanding the flow from downstream to terminal products and from manufacturing to testing [1][30] - The semiconductor industry is characterized by cycles, including long-term innovation cycles, capacity expansion cycles, and short-term inventory cycles, which are influenced by product launches and market demand [1][32][33] Domestic and International Trends - The trend towards domestic production in the semiconductor industry is progressing, with many segments achieving initial domestic production and beginning to internationalize [1][34]

Summary of Conference Call Records Industry Overview - The domestic computing power sector is benefiting from policy implementation and overseas market trends, with an overall positive outlook for the second half of the year. There is potential for significant price increases from late August to the end of the year as major companies provide KPI guidance and mid-year financial reports [1][2][19]. Key Companies and Recommendations - **Cambricon (寒武纪)**: Recommended as a key player in the domestic GPU and ASIC segments. If the n+2 expansion proceeds smoothly by the second half of 2025, the expected capacity could reach over 600 billion, supporting a substantial market valuation [1][5]. - **New Yuan Co. (新元股份)**: Still presents investment opportunities in the second half of the year, particularly due to its collaboration with Alibaba and potential overperformance in film progress, which could lead to a strong stock price reaction [1][8]. - **Switch Chip Sector**: Expected to grow significantly, with domestic companies like Shenghe Communication (盛和通信) likely to see major catalysts. The market for non-NVIDIA computing chips is projected to reach approximately $3.8 billion by 2026 [1][10][11]. - **ZTE and Hua Hong Semiconductor**: Both companies have positive expectations for Q3 and Q4, with ZTE launching new technologies and products, indicating strategic importance [1][14]. Market Dynamics - The domestic computing power sector has shown significant market performance since June, driven by supply and demand changes and chip testing progress. The overall trend remains positive despite some fluctuations [2][19]. - The relationship between GPU and ASIC segments is characterized by shared growth opportunities, with both experiencing high demand and tight supply [6]. Financial Projections and Market Size - The market for PCIe switch chips is projected to grow significantly, with estimates suggesting a market size of approximately $3.8 billion by 2026, potentially doubling by 2027 as technology upgrades occur [11]. - **Yongxin Electronics (永新电子)**: Expected to benefit from advanced manufacturing and domestic GPU client integration, with projections indicating a potential market valuation of up to 400 billion if certain sales targets are met [18][21]. Investment Opportunities - The recent adjustment in the Hong Kong stock market provides a good opportunity for investors to reposition, particularly in companies like Hua Hong Semiconductor, which is expected to have significant future developments [15][16]. - The domestic advanced packaging sector, including companies like Yuxi Electronics (有锡电子) and Yongxin Electronics, is highlighted as having substantial growth potential due to their roles as suppliers to major GPU clients [20][21]. Conclusion - The domestic computing power sector is poised for growth, with several key players and segments showing promise. Investors are encouraged to focus on companies with strong strategic positions and potential catalysts for growth in the coming months [19].

Summary of North American Cloud Business Conference Call Industry Overview - The North American cloud industry is experiencing strong growth, driven by investments in new applications and a subscription model for both consumer (C-end) and enterprise (B-end) clients [1][2] - The cloud computing market is currently valued in the hundreds of billions, while the AI market is projected to reach a trillion-dollar potential [1][6] Key Financial Metrics - North American cloud business has shown robust growth over the past decade, with a compound annual growth rate (CAGR) of 18% in revenue and 23% in net profit [2] - Free cash flow CAGR over the past ten years is 18%, with capital expenditures increasing at a CAGR of 26% over the same period [2] Business Model - The business model includes Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS), focusing on software development for higher profit margins [4] - IaaS involves significant infrastructure investments, while PaaS and SaaS rely more on software development [4] AI Impact - AI is viewed as a more promising field than traditional cloud computing, with current revenue from AI expected to reach $40-50 billion by the end of the year, but still far from the trillion-dollar market potential [5][6] - AI revenue is anticipated to be generated from intelligent agents and application tools, empowering businesses and creating new revenue streams [5] Market Dynamics - The demand for AI computing power is experiencing explosive growth, particularly in inference computing, with major players like Google and Microsoft significantly increasing their capital expenditures [3][15] - The AI era is marked by a surge in participants, with the top four companies now holding only 60-70% of the market share, compared to 80-90% during the cloud computing era [16] Equipment Lifecycle and Investment - Data center IT equipment typically has a high depreciation rate, requiring replacement every 2-3 years to maintain efficiency and performance [10][11] - Frequent equipment upgrades are necessary to meet the demands of Moore's Law, with new products often offering double the performance at a cost increase of only 30-50% [8][9] Emerging Players - New players like Oracle, Tesla, and Apple are emerging in the AI space, with Oracle's spending expected to reach significant levels by 2026 [17][18] - These companies are not only participating in capital expenditures but are also actively competing in the AI arms race, indicating a more complex and competitive market landscape [19] Future Market Implications - The presence of new entrants in the AI market will likely lead to a more diversified and competitive environment, with sustained growth driven by increased demand for inference computing [19][20] - The evolving market dynamics suggest that the AI landscape will continue to expand, with new players contributing to the overall growth and complexity of the industry [19]

Core Insights - Data centers are becoming the core engine driving global economic and social development, marking a new era in the semiconductor industry driven by AI, cloud computing, and large-scale infrastructure [1] - The demand for semiconductors in data centers is evolving from simple processors and memory to a complex ecosystem encompassing computing, storage, interconnect, and power supply [1] AI Surge: Arms Race in Data Centers - The explosion of artificial intelligence, particularly generative AI, is the most powerful catalyst for this transformation, with AI-related capital expenditures surpassing non-AI spending, accounting for nearly 75% of data center investments [3] - By 2025, AI-related investments are expected to exceed $450 billion, with AI servers rapidly increasing from a few percent of total computing servers in 2020 to over 10% by 2024 [3] - The global semiconductor market for data centers is projected to reach $493 billion by 2030, with data center semiconductors expected to account for over 50% of the total semiconductor market [3] GPU and ASIC Race - GPUs will continue to dominate due to the complexity and processing demands of AI workloads, with NVIDIA transforming from a traditional chip designer to a full-stack AI and data center solution provider [5] - Major cloud service providers are developing their own AI acceleration chips to compete with NVIDIA, intensifying competition in the AI chip sector [5] HBM Market Growth - The HBM market is experiencing explosive growth, expected to reach $3.816 billion by 2025, with a CAGR of 68.2% from 2025 to 2033 [6] - Key trends in the HBM market include increased bandwidth and capacity, energy efficiency, integration with AI accelerators, and the rise of standardized interfaces [6] Disruptive Technologies - Silicon photonics and co-packaged optics (CPO) are redefining data center performance and efficiency, with industry giants actively investing in this area [8] - The introduction of TFLN modulators is enhancing optical communication capabilities within data centers [9] Next-Generation Data Center Design - The shift to direct current (DC) power supply is becoming essential due to the rising power density demands of AI workloads, with modern AI racks requiring up to 600 kW [11] - Wide bandgap (WBG) semiconductor materials like GaN and SiC are crucial for high-frequency, high-voltage power conversion systems [12] - Liquid cooling technology is projected to grow at a CAGR of 14%, expected to exceed $61 billion by 2029, addressing the cooling challenges posed by high-density AI workloads [12] Advanced Thermal Management - Advanced cooling solutions, including direct chip liquid cooling and immersion cooling, are becoming necessary as traditional air cooling methods are insufficient for high-density AI workloads [13][14] - The industry is at a "thermal tipping point," necessitating fundamental adjustments in data center design to accommodate liquid cooling requirements [15] Future Outlook - The future of data centers will be characterized by increased heterogeneity, specialization, and energy efficiency, with a focus on advanced packaging technologies and comprehensive sensor systems [15]

Core Viewpoint - The article discusses the evolving landscape of artificial intelligence (AI) processing solutions, highlighting the need for companies to balance current performance with future adaptability in AI models and methods. Various processing units such as GPUs, ASICs, NPUs, and FPGAs are being utilized across different applications, from high-end smartphones to low-power edge devices [1][12]. Summary by Sections AI Processing Units - Companies are exploring a range of processing units for AI tasks, including GPUs, ASICs, NPUs, and DSPs, each with unique advantages and trade-offs in terms of power consumption, performance, flexibility, and cost [1][2]. - GPUs are favored in data centers for their scalability and flexibility, but their high power consumption limits their use in mobile devices [2]. - NPUs are optimized for AI tasks, offering low power and low latency, making them suitable for mobile and edge devices [2]. - ASICs provide the highest efficiency and performance for specific tasks but lack flexibility and have high development costs, making them ideal for large-scale, targeted deployments [3]. Custom Silicon - The trend towards custom silicon is growing, with major tech companies like NVIDIA, Microsoft, and Google investing in tailored chips to optimize performance for their specific software needs [4]. - Custom AI accelerators can provide significant advantages, but they require a robust ecosystem to support software development and deployment [4]. Flexibility and Adaptability - The rapid evolution of AI algorithms necessitates flexible hardware solutions that can adapt to new models and use cases, as traditional ASICs may struggle to keep pace with these changes [4][5]. - The need for adaptable architectures is emphasized, as AI capabilities may grow exponentially, putting pressure on decision-makers to choose the right processing solutions [4][5]. Role of DSPs and FPGAs - DSPs are increasingly being replaced or augmented by AI-specific processors, enhancing capabilities in areas like audio processing and motion detection [7]. - FPGAs are seen as a flexible alternative, allowing for algorithm updates without the need for complete hardware redesigns, thus combining the benefits of ASICs and general-purpose processors [8]. Edge Device Applications - Low-power edge devices are utilizing MCUs equipped with DSPs and NPUs to meet specific processing needs, differentiating them from high-performance mobile processors [10]. - The integration of AI capabilities into edge devices is becoming more prevalent, with companies developing specialized MCUs for machine learning and context-aware applications [10][11]. Conclusion - The edge computing landscape is characterized by a complex mix of specialized and general-purpose processors, with a trend towards customization and fine-tuning for specific workloads [12].

Core Insights - Data centers are becoming the core engine driving global economic and social development, marking a new era for the semiconductor industry, driven by AI, cloud computing, and large-scale infrastructure [2] - The demand for chips in data centers is evolving from simple processors and memory to a complex ecosystem encompassing computing, storage, interconnect, and power supply [2] AI Surge: The Arms Race in Data Centers - The explosion of artificial intelligence, particularly generative AI, is the strongest catalyst for this transformation, with AI-related capital expenditures surpassing non-AI spending, accounting for nearly 75% of data center investments [4] - By 2025, AI-related investments are expected to exceed $450 billion, with AI servers rapidly increasing from a few percent of total computing servers in 2020 to over 10% by 2024 [4] - Major tech giants are engaged in a fierce "computing power arms race," with companies like Microsoft, Google, and Meta investing hundreds of billions annually [4] - The data center semiconductor market is projected to expand significantly, reaching $493 billion by 2030, with data center semiconductors expected to account for over 50% of the total semiconductor market [4] Chip Dynamics: GPU and ASIC Race - GPUs will continue to dominate due to the increasing complexity and processing demands of AI workloads, with NVIDIA transforming from a traditional chip designer to a full-stack AI and data center solution provider [7] - Major cloud service providers are developing their own AI acceleration chips to compete with NVIDIA, intensifying competition in the AI chip sector [7] - High Bandwidth Memory (HBM) is becoming essential for AI and high-performance computing servers, with the HBM market expected to reach $3.816 billion by 2025, growing at a CAGR of 68.2% from 2025 to 2033 [8] Disruptive Technologies: Redefining Data Center Performance - Silicon photonics and Co-Packaged Optics (CPO) are key technologies addressing high-speed, low-power interconnect challenges in data centers [10] - The adoption of advanced packaging technologies, such as 3D stacking and chiplets, allows semiconductor manufacturers to create more powerful and flexible heterogeneous computing platforms [12] - The shift to direct current (DC) power supply is becoming essential due to the rising power density demands of modern AI workloads, with power requirements for AI racks expected to reach 50 kW by 2027 [13] Cooling Solutions: Liquid Cooling Technology - Liquid cooling technology is becoming a necessity for modern data centers, with the market projected to grow at a CAGR of 14%, exceeding $61 billion by 2029 [14] - Various types of liquid cooling methods, including Direct Chip Liquid Cooling (DTC) and immersion cooling, are being adopted to manage the heat generated by high-performance AI chips [15] - Advanced thermal management strategies, including software-driven dynamic thermal management and AI model optimization, are crucial for maximizing future data center efficiency [16] Future Outlook - The future of data centers will be characterized by increasing heterogeneity, specialization, and energy efficiency, with chip design evolving beyond traditional CPU/GPU categories [17] - Advanced packaging technologies and efficient power supply systems will play a critical role in shaping the next generation of green and intelligent data centers [17]

由追究起法律责任的权利各位投资人 各位领导 大家晚上好我是中国电子总统非常感谢大家到晚上这个时间来参加我们电话会然后今天的话 半导体这边也是大场估计也是因为几个上台言有外媒那边报道就是说可能国内未来会对这个部分国有企业国内一些企业要求说 那么前手对H20一个采购然后另外也有些市场团员就讲到国内的这边整个像SIMP这边的良率包括像汉武金那边的一个排单可能有所提升在多重的因素的结果之下今天的整个大盘表现非常好像班内这边的整个指数场服我们有两个多点然后其中像汉武金也是拉板了拉了一个20厘米的一个涨停然后像深谷和奥西科技也分别涨了9个点和5个点整体像新元的话整个市值已经逼近了前高马上就要到新高的阶段了然后奥迅的重回上行中道所以整体的一个国内的AS3Z这边表现其实非常旺盛 所以我们今天也借这个机会然后给各位领导汇报一下我们最近对新股份和奥迪科技的一个程度更新首先是新股份那么公司之前也发布了这个2025年的中期的应急预告那其中非常重要的是公司的一个在手订单其实有一个非常大的一个很显著的提升了就单Q2整体的在手订单是达到了30个亿那单Q2单个季度一个缓增相对于说Q1的一个这个订单增幅是达到6个亿那这个放在过去上市以来这个时 ...

Investment Rating - The report maintains an "Outperform" rating for the communication industry [1]. Core Insights - The upgrade of AI cluster network architecture is driving demand for optical modules, switches, network cards, and fiber optic cables due to the need for enhanced backend networking to support increased east-west traffic [2][12]. - The upcoming release of NVIDIA's GB300 is expected to significantly boost the demand for 1.6T optical modules, as the new architecture requires a higher number of upgraded optical modules compared to its predecessor [2][67]. - The trend towards self-developed ASICs by major players like Google, AWS, Meta, and Microsoft is anticipated to further increase the demand for optical modules and AEC connections [2][69]. Summary by Sections AI Cluster Network Architecture Upgrade - Traditional network architectures are inadequate for AI computing, primarily due to high latency and bandwidth limitations [11][12]. - The AI network architecture necessitates the addition of backend networks, which increases the demand for switches and optical modules [12][34]. NVIDIA H100 and NVL72 Cluster Network Architecture - The report details the network architecture for NVIDIA's H100 and NVL72 clusters, highlighting the differences in optical module requirements between the GB200 and GB300 models [47][66]. - The GB300 model requires a significant number of 1.6T dual-port optical modules, reflecting a shift in optical module demand as network speeds increase [67][68]. North American Leading CSP Network Architecture - The report discusses the network architecture of leading cloud service providers in North America, emphasizing the importance of efficient interconnectivity and the role of optical modules in supporting large-scale AI operations [74][78]. Investment Recommendations - The report recommends leading companies in the optical device and module sector, such as Zhongji Xuchuang, Xinyi Technology, and Tianfu Communication, while also suggesting attention to companies like Huilv Ecology and Zhaolong Interconnect in the high-speed copper cable segment [2][113].