Core Viewpoint - The PCB industry is experiencing significant growth driven by increasing demand for HDI products, particularly due to advancements in AI server technology and substantial capital investments from major tech companies [3][7]. Market Trends - The HDI market is projected to grow at a compound annual growth rate (CAGR) of 6.4% from 2024 to 2029, reaching a global market size of $17.037 billion by 2029 [3]. - Major companies like Google are increasing their capital expenditures, with a reported $22.446 billion for Q2 2025, marking a 30.5% quarter-over-quarter increase and a 70.2% year-over-year increase [3]. Company Performance - Shengyi Technology (生益电子) reported a 91% year-over-year increase in revenue for the first half of 2025, reaching 3.769 billion yuan, and a 452% increase in net profit to 531 million yuan [3]. Technological Developments - The PCB industry is evolving towards high-frequency, lightweight, and lead-free technologies, with CoWoP (Chip-on-Wafer-on-Panel) becoming a focal point for innovation [5][9]. - CoWoP technology requires Low-CTE (Coefficient of Thermal Expansion) materials to ensure compatibility with high-precision applications, leading to a focus on specialized materials and advanced manufacturing processes [9]. Key Drivers of Demand - The demand for PCBs is primarily driven by two segments: high-performance GPUs led by Nvidia and custom ASIC chips developed by cloud service providers like Google and Amazon [7][8]. - The shift towards custom ASICs is aimed at optimizing workloads and reducing reliance on external suppliers, which is expected to further increase PCB demand [8]. Advanced Packaging and Materials - CoWoP technology transforms PCBs from mere connectors to high-precision "inner substrates," necessitating advanced materials and manufacturing techniques [9]. - The introduction of HVLP (High-Voltage Low-Power) copper foil is essential for high-speed signal transmission, with surface roughness controlled to minimize signal loss [12][13]. Resin Innovations - The industry is moving away from traditional epoxy resins due to high dielectric loss, favoring advanced resin systems like PPO, PTFE, and CH, which offer better electrical performance [14].

Summary of Industrial Fulian Conference Call Company Overview - **Company**: Industrial Fulian - **Industry**: AI Computing and Server Manufacturing Key Points and Arguments Revenue and Market Capitalization Projections - Industrial Fulian raised its revenue forecast for 2026 to 50 billion yuan, with a market capitalization potentially reaching 750 billion yuan in 2025 and 1 trillion yuan in 2026, driven by significant increases in GB200/GB300 server shipments [1][4] - The GB200 server shipments are expected to triple quarter-over-quarter in Q3, with an annual projection of over 25,000 units, contributing approximately 6.5-6.6 billion yuan in net profit for every 10,000 cabinets sold [1][5] Impact of NVIDIA GPU Shipments - NVIDIA's GPU shipments are crucial for Industrial Fulian, with expectations of 8 million GPUs shipped in 2026, primarily for NVL cabinets, leading to double-digit growth in cabinet shipments and improved net profit margins due to scale effects [1][6] - An additional order of 300,000 H20 chips from NVIDIA is anticipated to enhance Industrial Fulian's performance, as the company is a supplier for B200 and other special modules [1][9] Liquid Cooling Technology - The development of liquid cooling technology is expected to improve both gross and net profit margins, with a dedicated subsidiary focusing on this area [1][10] AI Server Business Growth - The AI server business is projected to have high growth potential in 2025, with general servers benefiting from AI applications, maintaining double-digit growth from 2025 to 2026 [1][10] ASIC Server Development - Industrial Fulian aims to introduce Tranium ASIC servers from Amazon and plans to incorporate Google ASIC servers by 2026, which will enhance its ASIC server product line share [2][11] Stargate Project - The Stargate project is progressing well, with the first station operational and a 4.5GW agreement signed with Oracle. The project is expected to involve a total investment of 500 billion USD over five years, with a peak CAPEX of 150 billion USD, providing substantial revenue opportunities for Industrial Fulian [3][12] Capital Expenditure Plans - The company plans to increase capital expenditures by over 20% in 2025, with 60-70% allocated for capacity expansion, reflecting confidence in demand [3][13] Long-term Benefits in AI Computing - Industrial Fulian is positioned for long-term benefits in the AI computing sector due to the increasing demand for computing power across various applications, including AI and AGI, with significant growth potential in GPU deployment [1][15][16] - The company has strong ties with core customers, directly supplying NVIDIA modules and potentially entering the ASIC supply chains of Google and Amazon, which enhances its competitive advantage [1][16] Additional Important Insights - The server market is evolving with a focus on optimizing network and storage capabilities, which is critical for enhancing overall computing power [1][15] - The competitive landscape is expected to intensify, with macroeconomic risks and potential underperformance in the AI industry being notable concerns [1][16]

Core Viewpoint - The article discusses the architecture and networking components of the GB200 server, focusing on the use of copper and optical connections, and highlights the flexibility and cost considerations in the design choices made by different customers [1][2]. Frontend Networking - The frontend networking in the GB200 architecture serves as the main channel for external data exchange, connecting to the internet and cluster management tools [1]. - Each GPU typically receives a bandwidth of 25-50Gb/s, with total frontend network bandwidth for the HGX H100 server ranging from 200-400Gb/s, while GB200 can reach 200-800Gb/s depending on configuration [2]. - Nvidia's reference design for frontend networking may be over-provisioned, leading to higher costs for customers who may not need such high bandwidth [2][4]. Backend Networking - The backend networking supports GPU-to-GPU communication across large-scale clusters, focusing on internal computational collaboration [5]. - Various switch options are available for the backend network, with initial shipments using ConnectX-7 cards and future upgrades planned for ConnectX-8 [6][10]. - Long-distance interconnections primarily utilize optical cables due to the limitations of copper cables over longer distances [6]. Accelerator Interconnect - The accelerator interconnect is designed for high-speed communication between GPUs, significantly impacting communication efficiency and system scalability [13]. - The GB200's NVLink interconnect has evolved from the HGX H100, requiring external connections due to the separation of NVSwitches and GPUs across different trays [14]. - Different configurations (NVL72, NVL36x2, NVL576) balance communication efficiency and scalability, with NVL72 being optimal for low-latency scenarios [15]. Out of Band Networking - The out-of-band networking is dedicated to device management and monitoring, focusing on system maintenance rather than data transmission [20]. - It connects various IT devices through baseboard management controllers (BMC), allowing for remote management and monitoring of system health [21]. Cost Analysis of MPO Connectors - The article estimates the value of MPO connectors in the GB200 server, indicating that the cost per GPU can vary significantly based on network architecture and optical module usage [22][23]. - In a two-layer network architecture, the MPO value per GPU is approximately $128, while in a three-layer architecture, it can rise to $192 [24]. - As data center transmission rates increase, the demand for high-speed optical modules and corresponding MPO connectors is expected to grow, impacting overall costs [25].

Investment Rating - The industry rating is "Outperform the Market" (maintained rating) [6] Core Insights - The report highlights the launch of the GB300 AI server and the anticipated growth in demand for AI infrastructure, driven by the GB200 and GB300 servers [2][10] - The foldable iPhone has entered the P1 prototype stage, which is expected to benefit the foldable screen supply chain, with an early shipment estimate of 7 million units [1][14] - The AI-PCB market is experiencing growth due to increased demand from NVIDIA AI servers and ASIC requirements [3][32] Summary by Sections Section 1: Market Trends - The foldable iPhone project by Apple is in the P1 development stage, with a potential market rebound for foldable smartphones expected in 2026 [1][14] - The GB300 AI server, launched by NVIDIA, is set to significantly enhance AI inference capabilities, with a 50-fold increase in output and a 5-fold increase in throughput [2][20] - Marvell predicts that by 2028, over half of the $500 billion data center chip expenditure will be allocated to AI acceleration, with a CAGR of 35% for the custom ASIC market [2][24] Section 2: AI-PCB Growth - The demand for AI-PCB is driven by the requirements of NVIDIA's AI servers and the growth of ASIC technology, leading to a rapid increase in the AI PCB industry's value [3][32] - The GB200 NVL72 architecture from NVIDIA raises the standards for PCB, necessitating higher layer counts and advanced materials [3][35] - The ASIC development is pushing PCB demand, with the value of AI PCBs expected to quadruple in the coming years due to increased complexity and performance requirements [3][39] Section 3: Panel Industry Overview - The overall demand for panels is slowing, with a slight decrease in prices observed in July, particularly for television panels [4][41] - TCL Technology reported a significant increase in net profit for its semiconductor display business, while facing challenges in its solar energy segment [4][46] Section 4: Investment Recommendations - The report suggests focusing on various companies within the consumer electronics sector, including industrial and electronic component manufacturers, as well as companies involved in the foldable screen supply chain [5]

Investment Rating - The report maintains a "Buy" rating for the company with a target price of RMB 35.60, up from the previous RMB 30.00 [4][6]. Core Insights - The company is expected to achieve a net profit attributable to shareholders of RMB 119.58 to 121.58 billion for the first half of 2025, representing a year-on-year growth of 36.84% to 39.12%. The second quarter of 2025 is projected to see a net profit of RMB 67.27 to 69.27 billion, with a year-on-year increase of 47.72% to 52.11% [1][4]. - The growth is primarily driven by the accelerated shipment of the GB200 server and the ramp-up of 800G switches, indicating a strong performance in the second half of 2025 [1][2]. Summary by Sections Cloud Computing Business - In Q2 2025, the company's cloud computing revenue grew significantly, increasing by over 50% year-on-year. AI server revenue surged by more than 60%, while revenue from cloud service provider servers more than doubled [2]. - The growth is attributed to the large-scale shipment of the GB200 series AI chips/servers and increasing demand for AI cabinet products from cloud service providers [2]. Communication and Network Equipment Business - The communication and mobile network equipment segment saw steady growth, particularly in the smartphone metal frame business due to strong sales of high-end and budget AI models [3]. - The revenue from 800G switches reached three times the total revenue for 2024, driven by rising AI demand and an improving product mix [3]. Financial Projections - The report projects the company's net profit attributable to shareholders for 2025 to be RMB 303 billion, with expected growth rates of 30.3%, 24.1%, and 20.1% for 2025, 2026, and 2027 respectively [4][10]. - The earnings per share (EPS) are forecasted to be RMB 1.52, RMB 1.89, and RMB 2.27 for 2025, 2026, and 2027 respectively [10].

Summary of Conference Call Records Industry Overview - The conference call primarily discusses the PCB (Printed Circuit Board) industry, focusing on AI server demand and related materials such as copper foil and electronic cloth [1][2]. Key Points and Arguments AI Server Demand - The GB200 server has entered mass production, with an expected shipment of 20,000 to 25,000 cabinets this year. In the first half of the year, 12,000 cabinets were shipped, with a forecast of 13,000 cabinets for the second half. The price fluctuates between 35,000 to 40,000 yuan per square meter [1][3]. - The GB300 solution is set to adopt HDR high-layer designs, with sample shipments already sent out. Production is expected to start gradually in August, with an estimated shipment of 10,000 to 15,000 cabinets this year [1][5]. Supplier Landscape - Major suppliers for SVD high-layer boards include Huadian (40%), TTM (20%), Shenghong (20%), and Youde Maike (10%). The GB200 is expected to reach 25,000 cabinets this year and over 15,000 cabinets next year [2][4]. Material Challenges - The LDK third-generation quartz cloth faces challenges in drilling processes due to its high hardness, which complicates laser drilling and can lead to larger hole diameters and resin tearing. Improvements are needed in fiber cloth hardness and light absorption rates [1][6][7]. - New material development faces technical hurdles, particularly with Q cloth, which poses ion migration risks in small hole spacing designs. Solutions include optimizing product designs and collaboration between material suppliers and PCB manufacturers [1][8]. Equipment and Capacity - Core equipment, such as laser processing machines, has long lead times (12-15 months) due to high market demand, which limits capacity expansion. However, existing factories still maintain testing capabilities [2][9]. Market Trends - The domestic market shows increasing demand for high-end materials, with companies like Huawei leading the way. However, the overall impact has yet to be fully realized. Shengyi has captured the largest share in the BT storage substrate market and is collaborating with Huawei on advanced technologies [2][18]. - The iPhone-related server shipments are projected to reach 10 million units in 2025, with AR-related products accounting for 25%-30% of that volume. This percentage is expected to increase by 15-20 percentage points annually [2][12]. Future Outlook - The AR supply chain demand has exceeded expectations in the first half of 2025, while the smartphone and consumer electronics market saw a 20% decline in sales in the second quarter of 2025, leading to a drop in related substrate orders [2][17]. - Shengyi has opportunities to capture market share in the NV sector, leveraging its experience in high-layer substrates and connections within the NV supply chain [2][19]. Additional Important Information - The glass substrate solution is not yet fully defined, facing processing difficulties and yield issues when combined with ABF. Traditional organic substrates will continue to dominate in the short term [2][13]. - Domestic companies like Shengyi and Nanya are expanding their production capabilities, with Shengyi aiming to increase its market share in high-end materials [2][14][18].

Core Viewpoint - The financial analysis of Oracle's AI data center indicates that despite significant revenue, the operation is projected to incur substantial losses over five years, totaling approximately $10 billion [1][10]. Revenue - The average annual revenue over five years is projected to be $9,041 million, totaling $45 billion [3]. Hosting Cost - Hosting costs, which Oracle pays to data center service providers for GPU server placement, are expected to rise annually due to inflation and market conditions [4]. Electricity Cost - Electricity costs, a fixed expense associated with high-load GPU operations, are also anticipated to increase slightly each year [5]. Gross Profit - The largest cost in the financial model is server depreciation, estimated at $3.3 billion annually, leading to a total asset depreciation to zero within seven years [7]. Operating Profit - Operating profit is significantly impacted by interest expenses, which are expected to total $3.6 billion over the first four years, with a notable reduction in the final year [8]. Contribution Profit - After accounting for taxes, the annual contribution profit is projected to be around $2.5 billion, resulting in a total of $12.5 billion over five years [10].

Investment Rating - The report maintains an "Overweight" rating for the electronics industry, expecting the industry index to outperform the market index by more than 10% in the next six months [33]. Core Insights - The Taiwanese ODM manufacturers have seen significant year-on-year revenue growth, primarily driven by the accelerated shipment of GB200 servers. The global AI server shipment is projected to reach 1.811 million units in 2025, representing a year-on-year growth of 26.29%, with high-end AI servers equipped with HBM expected to grow over 40% [2][29]. - The SW electronics sector's PE TTM (excluding negative values) is 42.23 times, which is at the 73.33% percentile for the past five years and 62.75% for the past ten years, indicating a relatively high valuation [2][14]. Market Review and Valuation - The SW electronics sector increased by 0.67% over the past two weeks (05/30-06/12), underperforming the CSI 300 index by 0.20 percentage points, ranking 16th among the Shenwan industries. In June, the sector rose by 2.57%, outperforming the CSI 300 index by 1.21 percentage points, ranking 14th. However, the sector has declined by 2.01% year-to-date, underperforming the CSI 300 index by 0.92 percentage points, ranking 23rd [9][10]. Industry News - Huawei launched the Pura 80 series, featuring the eighth-generation ISP and a one-inch RYYB ultra-large sensor. NVIDIA's CEO announced plans to establish over 20 large AI factories in Europe during the VivaTech conference. OpenAI reported an annual recurring revenue of $10 billion, up from $5.5 billion last year, driven by its consumer products and API revenue [19][21]. - The average daily token usage of the Doubao large model exceeded 16.4 trillion, a 137-fold increase since its launch last year, with a 46.4% market share in China's public cloud model service calls [19][21]. Industry Data - In Q1 2025, global smartphone shipments reached 305 million units, a year-on-year increase of 1.53%. In April 2025, China's smartphone shipments were 22.29 million units, down 1.65% year-on-year [21][25]. - The prices of liquid crystal panels remained stable in May 2025, with 32-inch, 43-inch, 50-inch, 55-inch, and 65-inch panels priced at $39, $65, $104, $130, and $178 per piece, respectively [25][26]. Weekly Perspective - Major Taiwanese ODM manufacturers, including Hon Hai, Quanta, and Wistron, have experienced significant revenue growth, benefiting from the accelerated shipment of GB200 servers. The report suggests focusing on the overseas computing power supply chain, particularly the PCB/CCL sector with high-end capacity and consistent performance [29][30].