Core Viewpoint - The company is simultaneously advancing its SST and Panama power supply projects to meet diverse customer power supply needs [1] Group 1: Technical Development - The company is conducting full-stack research and development across multiple voltage conversion paths to achieve comprehensive coverage from high voltage to core chip low voltage power supply [1] - Different customers are at various technological stages and may require different overall architecture technical solutions, necessitating customized development as a component supplier [1] - The overall progress of the work is steady and ongoing [1]

Core Insights - The article highlights the significant growth and future projections of NVIDIA's GPU series, particularly the Hopper, Blackwell, and Rubin series, indicating a strong revenue potential in the upcoming quarters [1][3]. GPU Shipment and Revenue - NVIDIA's Hopper series GPUs have shipped 4 million units, generating $100 billion in revenue [1] - The Blackwell series GPUs are expected to ship 20 million units over the next five quarters, with projected revenue of approximately $500 billion, marking a fivefold increase compared to the H series [3]. Optical Module Demand - Excluding the already shipped 6 million units, an additional 14 million optical modules are expected to be shipped in the next five quarters, averaging about 2.8 million units per quarter [4] - The demand for 1.6T optical modules is projected to exceed 34 million units, significantly surpassing previous market expectations of 20 million units by 2026 [6]. Pricing and Supply Dynamics - The anticipated supply shortage of 1.6T silicon optical chips is estimated at 10-20%, which may lead to order delays or shifts towards 800G optical modules [6] - The expected price for 1.6T optical modules has been adjusted to $1,000-$1,100 per unit, up from the previous forecast of $800-$900, potentially increasing revenue contributions by 15-20% [6]. Industry Valuation Concerns - The low valuation of the optical module industry is attributed to several factors, including reliance on North American markets, technological risks from new innovations, low manufacturing barriers, and uncertainty in long-term demand [8]. PCB Developments - The Rubin architecture will utilize M9 materials for its PCB designs, enhancing the value of the boards compared to previous generations [9][10]. - The total value of the PCB in the VR200 NVL144 CPX computing tray is projected to exceed $3,000, significantly higher than earlier models [11]. SST Implementation - The demand for SST (Smart Supply Technology) solutions is expected to accelerate, with major companies like Amazon seeking to test SST samples for potential integration in future products [16][17]. - The optimistic outlook for Rubin's shipment pace may enhance the valuation of the SST industry and the performance of related companies in the coming years [17].

Group 1: AI Market Growth - The global AI market is expected to grow at a CAGR of over 35% from 2023 to 2030, driven by the demand for intelligent computing [2][9] - The domestic market for AI computing power is projected to reach 26.7 billion by 2027, presenting significant opportunities for domestic manufacturers [3] Group 2: Semiconductor Testing Equipment - The importance of semiconductor back-end testing is increasing due to the complex structure of AI chips and HBM, necessitating more frequent and stringent testing [4][5] - There is substantial room for improvement in the domestic semiconductor back-end testing equipment market, indicating potential growth for local manufacturers [6] Group 3: Power Equipment - The demand for intelligent computing driven by AI is highlighting the need for advanced power distribution solutions in data centers, with SST solutions being a preferred choice [7][9] - Key players in the transformer sector, particularly those with expertise in high-frequency transformers, are recommended for investment as they are well-positioned to benefit from this trend [7][8]

Investment Rating - The report suggests a positive investment outlook for the SST (Solid State Transformer) solution, highlighting its comprehensive advantages in the data center sector [4][70]. Core Insights - The global AI market is projected to grow from 1,187.9 billion RMB in 2023 to 11,455.4 billion RMB by 2030, with a CAGR exceeding 35% [5][14]. - The SST solution is particularly suitable for microgrid DC power scenarios, enabling direct conversion from grid power to 800V DC supply, addressing various challenges in data center power supply [5][47]. - The report emphasizes the importance of high-frequency transformers and power electronic devices in the SST solution, recommending key companies in these sectors for investment [5][70]. Summary by Sections 1. Industry Status and Challenges - The demand for full DC systems is increasing due to the rising proportion of DC devices in data centers, necessitating a compatible full DC power supply architecture [21]. - Traditional 54V power supply systems are reaching their limits, prompting the need for higher voltage levels, such as the 800V HVDC architecture proposed by leading companies [28]. - The area occupied by power distribution systems is increasing as power density in data centers rises, with projections indicating that power distribution systems may occupy equal space to IT equipment in high-density setups [30] [31]. 2. Advantages of SST Solution - The SST solution offers modularity and high efficiency, with the ability to directly convert high voltage AC to low voltage DC, significantly reducing space requirements and improving energy efficiency [44][63]. - SST technology utilizes advanced semiconductor devices and high-frequency transformers, allowing for active control of power quality and rapid fault detection [55][57]. - The SST solution has already been implemented in benchmark data center projects, demonstrating its potential for widespread adoption in the industry [65]. 3. Investment Recommendations - The report recommends focusing on companies involved in high-frequency transformers, power electronic devices, and advanced magnetic materials, as they are expected to benefit from the growth of the SST market [70].

Core Viewpoint - The GTC conference highlighted the transition from Power shelf to Power rack as the next-generation power solution for AI data centers, indicating a significant upgrade in the importance of power suppliers in the industry [2][4]. Group 1: Power Rack Development - The Power rack integrates multiple components such as Power shelf, BBU shelf, Super cap shelf, and PDU, enhancing the overall value and competitive barriers in the AI power supply industry [2][3][6]. - The shift to Power rack architecture addresses the increasing power demands of advanced computing nodes while optimizing space utilization within data center racks [4][6]. Group 2: SST Technology - SST (Solid State Transformer) is identified as the future power supply technology, capable of converting high voltages (10-34.5KV) to 800V DC, facilitating modular design and voltage adaptability across different regions [3][12]. - The implementation of SST technology is expected to reduce power loss significantly by increasing voltage levels, thus minimizing current and associated losses [15][19]. Group 3: Super Capacitor Benefits - The use of super capacitors can significantly reduce power supply volatility, with Delta's data showing a reduction in supply fluctuation from 73% to 6% [20][24]. - Different super capacitor technologies exhibit varying characteristics, with EDLC offering longer cycle life and LiC providing higher energy density, which can enhance the dynamic load capacity of GPUs in AI data centers [21][22]. Group 4: Future Roadmap - NVIDIA anticipates the gradual application of the Rubin and Feyman architectures between 2026 and 2028, which will further accelerate the adoption of Power rack solutions [11][12]. - Google's roadmap for data center power architecture evolution indicates a transition towards higher voltage levels and modular designs, aligning with the industry's shift towards SST solutions [13][15].