Group 1 - The State Intellectual Property Office of China has granted a patent to Kehua Data Co., Ltd. for a "liquid cooling system control method and liquid cooling system," with the authorization announcement number CN117915634B and an application date of January 2024 [1] - Kehua Data Co., Ltd., established in 1999 and located in Xiamen, primarily engages in the manufacturing of electrical machinery and equipment, with a registered capital of 515.414041 million RMB [1] - The company has invested in 47 enterprises, participated in 3,874 bidding projects, holds 130 trademark registrations, and has 1,399 patents, along with 36 administrative licenses [1] Group 2 - Xiamen Huairuisheng Intelligent Technology Co., Ltd., founded in 2013 and also based in Xiamen, focuses on software and information technology services, with a registered capital of 100 million RMB [1] - The company has invested in 1 enterprise, participated in 20 bidding projects, holds 84 patents, and possesses 20 administrative licenses [1]

Core Viewpoint - The National Grid's "14th Five-Year Plan" investment is set to reach 4 trillion yuan, marking a 40% increase from the previous plan, focusing on green transformation and new power system construction [2][3]. Group 1: Investment and Development Plans - The National Grid aims to enhance its investment efforts to support domestic demand and stabilize growth, with a focus on aligning with national strategies and promoting social investment [3]. - By 2030, the "West-to-East Power Transmission" capacity is expected to exceed 420 million kilowatts, with an additional inter-provincial power exchange capacity of around 40 million kilowatts, supporting a 30% share of renewable energy generation [3]. Group 2: Power Supply Expectations - During the "14th Five-Year Plan," the clean energy installed capacity, represented by wind and solar power, is expected to continue growing, transitioning the power supply structure from fossil fuels to reliable renewable energy support [4]. - The power system is evolving from the traditional "source-grid-load" model to a "source-grid-load-storage" model, incorporating various new technologies [4]. Group 3: AI and Power Supply - The power supply is critical for AI development, with the U.S. facing a power shortage due to AI data centers, while China is projected to have three times the power generation capacity of the U.S. by 2026, supporting high-energy AI data centers [4]. - The increase in power supply is seen as a challenge that is often underestimated [4]. Group 4: Investment Opportunities - Analysts suggest focusing on the transformer sector, particularly in the context of technological innovation and demand growth, with specific recommendations for companies like Siyi Electric and Huaming Equipment [5]. - Solid-state transformers (SST) are on the verge of commercial explosion, with prototype validation expected in 2026 and commercial rollout anticipated in 2027, involving companies such as Sifang Co. and China XD Electric [5]. - The anticipated increase in investment during the "14th Five-Year Plan" presents opportunities in the ultra-high voltage sector and new standard electric meters, with companies like Pinggao Electric and XJ Electric highlighted for their potential [5].

Core Insights - The fundamental driver of AI power supply solutions is the continuous improvement in single-chip and single-cabinet power levels, with companies like NVIDIA leading the charge in upgrading their AI chips [1] - The power of AI data centers in North America is projected to reach 71GW by 2028, driven by advancements in multi-chip designs such as NVL72 from NVIDIA and Superpod from Google [1] Group 1: Power Supply Evolution - The main directions for AIDC power supply iteration are high power, high voltage, and direct current, with NVIDIA's white paper outlining a clear path from AC to 800V DC solutions [2] - The OCP organization has established ±400V power supply standards, aiming for MW-level cabinet power through direct current supply [2] Group 2: Technological Changes and Growth - The evolution of power supply solutions has led to the development of integrated devices like HVDC Sidecar and SST hosts, which can perform multiple functions such as voltage reduction and power quality improvement [3] - Key challenges in R&D include power conversion modules and high-frequency isolation transformers, with third-generation wide bandgap semiconductors like SiC and GaN being crucial for achieving high voltage and efficiency [3] Group 3: Investment Opportunities - Four key areas for investment consideration include AIDC power hosts (PSU, HVDC, SST), with companies like Sungrow Power (300274.SZ) and Megmeet (002851.SZ) being notable mentions [4] - Energy storage solutions at the station level are highlighted, with companies such as Sungrow Power and Fluence Energy (FLNC.US) being potential investment targets [4] - Core incremental components like solid-state circuit breakers and electronic fuses are also identified, with companies like Liyang Technology (002706.SZ) and Vicor Corporation (VICR.US) being recommended [4] - Third-generation semiconductors like GaN and SiC are emphasized as critical components for future developments [4]

Group 1 - DeepSeek is set to release its flagship AI model, DeepSeek V4, in February, which reportedly surpasses current top models in programming capabilities [1] - The core innovation of V4 is the Engram module, which separates knowledge storage from logical reasoning, allowing for efficient retrieval of static knowledge [2][3] - The Engram module is expected to reduce the reliance on high-cost GPU memory (HBM) by migrating 20%-25% of static knowledge parameters to main memory (DRAM), significantly altering the model's storage requirements [3] Group 2 - AI programming is a key focus for major companies, with DeepSeek's advancements potentially enhancing the usage of domestic integrated development environments (IDEs) and benefiting low-code platforms [4] - The upcoming V4 model may improve the cost-effectiveness of AI applications and could support domestic chip architectures, which would accelerate the development of the domestic AI industry [5] - Historical performance of DeepSeek's previous model, R1, saw significant stock price increases, indicating strong market interest in its AI technologies [6] Group 3 - Relevant companies in the SSD storage sector include Jiangbolong, Demingli, and Baiwei Storage, while application vendors include Hehe Information, Wanxing Technology, and others [9] - Companies involved in computing infrastructure include Cambricon, Haiguang Information, and others, indicating a broad ecosystem supporting DeepSeek's advancements [9]

Summary of Key Points from the Conference Call Transcript Industry Overview - **Industry**: China Data Center Equipment - **Key Focus**: Power and cooling upgrades, particularly related to NVIDIA's Vera Rubin platform launch in 2026 [2][3] Core Insights and Arguments - **NVIDIA's Rubin Platform**: - The Rubin platform is set to double the rack power compared to the previous Blackwell platform, moving from 80% to 100% liquid cooling [2] - Rubin is currently in full production [2] - **VR200 NVL72 Specifications**: - Delivers approximately 3.5x to 5x the training/inference AI computing power compared to GB300 NVL72, significantly increasing rack-level power demand [3] - Upgrades power shelves to a 3*3U 110 kW configuration, compared to the GB300's 8*1U 33 kW shelves [3] - Introduces a 3+1 redundancy design for power shelves [3] - **Future Developments**: - Anticipation of transitioning to the next-generation Kyber rack design, which will support 800V HVDC to meet rising power requirements driven by AI compute scaling [3] - Potential launch of Rubin Ultra in 2027 could further unleash HVDC demand [3] Investment Opportunities - **Stock Picks**: - **Kstar and Kehua**: Expected to benefit from stronger UPS demand and potential upgrades to HVDC and SST [4] - Kstar is anticipated to have strong order intake from US hyperscalers due to strategic partnerships [4] - Kehua's domestic GPU ramp-up and H200 shipment could enhance order intake from domestic hyperscalers [4] Risks and Valuation - **Downside Risks for Data Center Equipment Sector**: - Slower-than-expected AI data center capacity growth [6] - Slower penetration of high-power density products [6] - Challenges in gaining market share in the overseas AIDC equipment supply chain [6] - **Kehua's Price Target and Risks**: - Price target based on DCF methodology; downside risks include slower IDC capacity expansion and lower overseas shipment expectations [7] - Upside risks include faster IDC capacity expansion and stronger relationships with hyperscalers [7] - **Kstar's Price Target and Risks**: - Similar DCF methodology for price target; downside risks include slower IDC capacity expansion and new entrants in the market [8] - Upside risks include faster IDC capacity expansion and higher overseas shipments [8] Additional Important Information - **Analyst Team**: The report was prepared by UBS Securities Asia Limited, with analysts Yishu Yan, Anna Yuan, and Ken Liu involved [5] - **Valuation Methodology**: The report emphasizes the use of DCF methodology for price targets and highlights the importance of understanding risks before making investment decisions [7][8] This summary encapsulates the critical insights and data from the conference call, focusing on the developments in the China data center equipment industry, particularly regarding NVIDIA's advancements and the implications for investment opportunities in Kstar and Kehua.

Summary of Liquid Cooling Industry Research Industry Overview - The liquid cooling industry is driven by the increasing power consumption of GPUs, particularly when GPU power exceeds 1,000 watts, making traditional air cooling insufficient [1][2] - Liquid cooling technology aligns with carbon neutrality goals due to its energy efficiency and superior heat dissipation capabilities [1][2] Key Points Current Technology and Market Trends - Cold plate liquid cooling is the dominant technology, accounting for over 90% of the market, with single-phase liquid cooling being the primary form [1][5] - Despite the potential of microchannel and phase change technologies, traditional single-phase non-direct contact cold plate liquid cooling remains the leading solution in the short term [1][5] - The application of liquid cooling in data centers is increasing, especially for high-power GPU cabinets, despite initial costs being 10%-20% higher than air cooling [1][6] Cost and Efficiency - Liquid cooling systems are more energy-efficient, reducing air conditioning energy consumption and lowering Power Usage Effectiveness (PUE) [2][6] - The cost of liquid cooling systems ranges from 3,000 to 3,500 yuan per kilowatt, compared to 2,500 to 3,000 yuan for air cooling, but the performance advantages make liquid cooling increasingly attractive [6] Market Dynamics - The power equipment configuration ratio in data centers is shifting from 70% IT load and 30% infrastructure in air cooling to 80% IT load and 20% infrastructure in liquid cooling [7] - Major suppliers in the liquid cooling market include companies like Vertiv, Delta, Baode, and Schneider, with Vertiv holding a significant share due to its collaboration with NVIDIA [3][13] Future Projections - The large-scale application of liquid cooling technology is expected to surge from 2024 to 2025, particularly in overseas markets, driven by the deployment of high-power GPUs [4] - The domestic market is anticipated to follow suit by 2026, contingent on policy changes and access to high-performance chips [4] Challenges and Considerations - Liquid cooling systems require careful design to account for IT load and CPU ratios, often necessitating redundancy to ensure stability under extreme conditions [7][9] - The integration of power and cooling systems is a growing trend, with major players acquiring companies to enhance their competitive edge and provide comprehensive solutions [21][22] Market Potential - The liquid cooling market has significant growth potential, with estimates suggesting a market size reaching billions if 40,000 units are deployed [20] - The trend towards prefabricated solutions in power and cooling systems is expected to reshape the competitive landscape in the industry [22] Additional Insights - The cooling liquid prices have decreased significantly, further narrowing the cost gap between liquid and air cooling systems [6] - The future of liquid cooling technology may involve the use of smaller refrigeration units to achieve lower temperature cooling, indicating a shift towards more advanced cooling solutions [12]

Core Viewpoint - The company has established a close partnership with Tencent Cloud in the data center sector, providing comprehensive lifecycle services tailored to Tencent Cloud's specific needs [1] Group 1: Partnership and Collaboration - The company has designed, built, and operated four customized data centers in Zhangjiakou, Guangzhou, and Qingyuan for Tencent Cloud, offering services from consulting and planning to operation management and value-added services [1] - The company collaborates with Tencent Cloud on an elastic resource pool model, integrating Tencent Cloud resources into data centers in Shanghai and Guangdong to enhance user convenience [1] Group 2: Product Application - The company's products, including UPS, high-voltage direct current, precision distribution, dynamic environment monitoring, MDC integration products, and TB container data centers, are utilized in multiple projects within Tencent's data centers [1] Group 3: Future Outlook - The company aims to continue collaborating with Tencent Cloud and other industry partners to create value, enhance product and service quality, and contribute to the innovative development across various industries [1]

Group 1 - The core point of the article is that Kehua Data (002335) has approximately 108,000 shareholders as of January 9, 2026 [1]

Investment Rating - The industry investment rating is positive, indicating an overweight outlook for the AIDC sector, suggesting it will outperform the overall market [8]. Core Insights - The AIDC industry is expected to experience significant breakthroughs in computing power, models, and applications in 2026, with domestic computing power transitioning from usable to highly effective [2]. - The demand from cloud vendors remains strong, with a notable increase in IDC orders expected to be fulfilled by mid-2026, indicating a robust growth trajectory for cloud computing infrastructure [2]. - The supply side faces challenges due to the scarcity of effective supply, as the government emphasizes orderly guidance and efficient construction of intelligent computing centers [2]. Summary by Sections Demand Side - Cloud vendors are expected to have high demand for IDC, with orders from Q1 2025 anticipated to be deployed by Q2-Q3 2026, necessitating significant capital investment [2]. - The new round of IDC bidding has commenced, with an accelerated delivery schedule expected [2]. Supply Side - The government is focusing on optimizing the layout of intelligent computing resources, which will create a balanced supply-demand situation in the long term [2]. - There is a noted expectation for supply of supporting equipment, such as UPS and HVDC systems, to meet the needs of AIDC construction [2]. Key Companies - Key players in the industry include leading IDC companies like Runze Technology and Dongyangguang, as well as emerging companies like Aofei Data and Guanghuan New Network [2]. - Companies providing essential supply and cooling solutions, such as Kehua Data and Zhongheng Electric, are also highlighted as important players [2].

26年1月10日下午，智谱创始人唐杰、千问技术负责人林俊旸、腾讯首席AI科学家姚顺雨等AI前沿工作 者AGI-Next前沿峰会上形成一定行业共识：在硬件支持、软硬件结合、Agent路径等多个维度上，国内 科技公司正在从全球AI技术的跟随者逐渐转变为推动者。 需求侧：云厂商需求高企 从必要性的角度看，25Q1前的IDC订单，预计到26Q2-Q3完成上架部署，而从下定到交付需要约6个月 缓冲期，这意味着云厂商25年底至26年初进行下一轮智算中心布局必要性很强，且节奏紧体量大。从实 际布局的情况看，海外算力与国产算力并行确定性高，预计将推动云厂商资本开支持续增长;新一轮IDC 招投标已启动，预计交付上架节奏显著加快。此外，云厂商有一定概率锁定更长时间维度内的优质区位 资产，以保障未来算力供给的可持续性。 申万宏源发布研报称，展望2026年，预计国内算力、模型、应用等多个层面均将有重大突破。算力侧， 除NV供应变化之外，国产算力预计从可用变为好用，并通过超节点方式实现更高密度算力集群组网。 模型侧，中美大模型差距在2025年逐步缩小，预计2026年追赶趋势将持续，并进一步达成性能与成本间 的更优平衡。应用侧，除25年 ...