Investment Rating - The report maintains an "Overweight" rating for the mechanical equipment industry [1] Core Insights - The GTC conference highlighted the increasing demand for computing power, particularly in AI inference, with significant advancements in NVIDIA's product architecture [1][2] - The introduction of the LPU cabinet architecture, designed for inference, emphasizes low latency and high throughput, which is expected to drive demand for PCB materials and liquid cooling solutions [2][4] - The report indicates a strong upward trend in PCB investments, with major companies planning significant capital expenditures to expand production capacity [4] Summary by Sections GTC Conference Highlights - NVIDIA's GTC conference showcased the necessity for future computing power and introduced new architectures such as Rubin and Groq [1] - The LPU architecture was a key highlight, featuring a design that integrates large SRAM directly on the chip, significantly reducing data access latency [2] PCB and Liquid Cooling Developments - PCB importance is increasing, with market space expected to expand due to rising investments from companies like Huadian and Pengding [4] - Liquid cooling is becoming essential as server architectures evolve, with a complete shift towards liquid cooling systems anticipated [4] Investment Recommendations - Recommended companies in PCB equipment include Dazhu CNC and Chipbond, while PCB materials recommendations include Zhongtung High-tech and Dingtai High-tech [5] - For liquid cooling, recommended companies include Yingweike and Hongsheng Shares, with additional companies suggested for consideration [5]

Investment Rating - The industry investment rating is "Positive" (maintained) [1] Core Insights - OpenClaw, an open-source AI agent framework, has gained significant popularity since its release, becoming the most popular open-source project on GitHub as of February 2026. It offers capabilities such as local-first operation, autonomous tool invocation, cross-application execution, multi-IM access, and continuous online operation, making it highly appealing to users [4] - The demand for inference computing power is expected to grow exponentially due to the characteristics of Agent AI, which include autonomous task execution and multi-agent concurrency. The daily token consumption in China surged from 100 billion in early 2024 to over 30 trillion by mid-2025, reflecting a growth of over 300 times in just one and a half years [5] - OpenClaw is set to transform AI from a conversational role to an execution role, enabling more intelligent personal AI assistants and digital employees. Major companies like Xiaomi and Huawei are adapting OpenClaw for various devices, indicating a shift in the capabilities of edge AI [5] Summary by Sections OpenClaw Development - OpenClaw was initially released on November 24, 2025, and saw a surge in adoption in February 2026, with major domestic companies launching products based on its code. Notable implementations include Tencent's WorkBuddy and Zhizhu's AutoClaw, which offer seamless integration with existing communication tools [4] Token Consumption and Demand - The introduction of Agent AI has led to a shift in token consumption patterns, with tasks now consuming tens of thousands to millions of tokens per execution. Predictions indicate that by 2030, the global active AI agents will reach 2.216 billion, with annual token consumption skyrocketing from 0.0005 Peta Tokens in 2025 to 15.2 million Peta Tokens [5] Investment Recommendations - The report suggests focusing on the AI inference computing power supply chain, including segments like chips, complete machines, liquid cooling, and power supplies. Recommended stocks include Haiguang Information, Lingyi Technology, and Dongshan Precision, among others [6]

Core Viewpoint - The recent developments in the computing power industry indicate a shift towards a more pragmatic phase, focusing on revenue and profit rather than merely expansion, as highlighted by OpenAI's adjustment in its investment strategy [1] Group 1: OpenAI's Investment Strategy - OpenAI has significantly reduced its computing power investment, which has sparked widespread discussion in the industry [1] - The adjustment is seen not as a budget cut but rather a change in expression, shifting from an "8-year broad infrastructure" approach to a "5-year computing power special" focus [1] Group 2: Industry Trends - The AI industry is not experiencing a bubble burst or a halt in development; instead, it is entering a more realistic stage that emphasizes commercial fundamentals [1] - North American cloud providers are expected to continue increasing capital expenditures, which will sustain high prosperity in the computing power and NVIDIA supply chain [1] Group 3: Evolving Needs in the Industry - As AI applications accelerate, the demand for computing power structure, AI large models, and business models within the supply chain will change [1] - Investment opportunities in AI inference computing power and domestic AI chips are highlighted as areas worth focusing on [1]

Group 1 - The semiconductor equipment ETF E Fund (159558) has seen significant inflows, with nearly 50 million shares net subscribed and a turnover rate of 5.51%, resulting in a transaction volume of 241 million yuan [1] - As of January 20, the latest scale of the semiconductor equipment ETF reached 4.351 billion yuan, with a total of 1.916 billion shares, both hitting record highs since its inception [1] - Over the past 16 days, the semiconductor equipment ETF has experienced continuous net inflows, with a peak single-day net inflow of 621 million yuan, totaling 2.366 billion yuan [1] Group 2 - The Greater Bay Area Compound Semiconductor Ecological Application Conference and Semiconductor Industry CEO Conference was held in Zhuhai High-tech Zone, where leaders expressed hopes for more companies to establish R&D centers, manufacturing bases, and regional headquarters in Zhuhai to create a globally influential compound semiconductor industry hub [1] - Western Securities anticipates sustained growth in demand for server CPUs due to upgrades in data center architecture and increasing AI inference computing power needs, with domestic server CPUs improving in performance and software compatibility, potentially increasing market share driven by demand and policy [1]

Investment Rating - The industry investment rating is "Overweight" [5] Core Views - The demand for server CPUs is increasing due to the upgrade of data center architectures and the continuous rise in AI inference computing power, leading to sustained growth in demand [2][3] - Intel and AMD are raising server CPU prices by 10%-15% to address supply-demand imbalances and ensure stable future supply, with their server CPU capacity for 2026 nearly sold out [1][2] - The general server market is recovering, with a projected global server shipment growth of over 9% year-on-year, driven by data center architecture upgrades and the replacement of existing server CPUs [1][2] Summary by Sections Section 1: Price Adjustments and Market Dynamics - Intel and AMD are increasing server CPU prices by 10%-15% due to supply-demand imbalances [1] - The global server shipment is expected to grow by over 9% year-on-year, influenced by the launch of new CPU products and data center upgrades [1][2] Section 2: AI Influence and Capital Expenditure - The rise of generative AI is driving an increase in AI server procurement, which is affecting the budget for general servers [2] - Cloud vendors are expanding capital expenditures to meet the growing demand for AI inference servers, with global AI server shipments projected to grow over 20% year-on-year by 2026 [2] Section 3: Domestic CPU Developments - Domestic next-generation server CPUs are accelerating deployment in various scenarios, with improvements in stability and compatibility [2][3] - Companies such as Loongson Technology, Haiguang Information, and China Great Wall are highlighted as key players in the domestic CPU market [3]

Investment Rating - The industry rating is "Outperform the Market," indicating that the overall return of the industry is expected to exceed the market benchmark index by more than 5% in the next 6 months [14]. Core Insights - Nvidia has formed a strategic partnership with AI chip startup Groq to enhance inference technology, with Groq's core team joining Nvidia while continuing to operate independently [3]. - The inference market is experiencing rapid expansion, with a significant increase in demand for inference computing power as the global AI industry transitions from model training to large-scale inference applications [4]. - Groq's technology, featuring LPU and SRAM architecture, optimizes inference performance, allowing for predictable, low-latency execution of large language models [5]. Summary by Sections Event Background - Nvidia and Groq have established a non-exclusive licensing agreement for inference technology, with Groq's leadership team integrating into Nvidia to advance technology upgrades and applications [3]. Industry Trends - The demand for inference computing power is growing rapidly, with Google processing 980 trillion tokens monthly as of July, doubling since May, and domestic models exceeding 50 trillion daily calls, a tenfold increase year-on-year [4]. Groq's Technical Value - Groq's LPU design focuses on speed and accuracy, utilizing static scheduling and deterministic execution to enhance performance in inference tasks. The second-generation LPU is manufactured using Samsung's 4nm process technology, improving speed and efficiency [5]. Investment Recommendations - The report suggests a positive outlook on the growth of inference computing demand and related supply chains, recommending attention to wafer fabrication and upstream equipment expansion in the domestic market, as well as growth in optical modules and cabinet assembly outsourcing in the overseas market [6].

Group 1 - Qiniu Cloud (七牛云) and Wuxiang Cloud Valley (五象云谷) have established a strategic partnership to scale AI inference computing power, targeting the trillion-dollar AI inference market [1][4] - Qiniu Cloud has three main product lines: Media Cloud, AI Cloud, and LinX Cloud, with a recent focus on AI inference computing products, marking a comprehensive layout from cloud to edge [1][3] - In the first half of 2025, Qiniu Cloud's AI-related revenue reached 184 million yuan, accounting for 22.2% of total revenue, primarily from AI inference services and computing resource leasing [3] Group 2 - The partnership aims to enhance the commercialization potential of Wuxiang Cloud Valley's computing resources, which has a total investment of 3.6 billion yuan and can provide up to 40,000 PetaFLOPS of intelligent computing power [4] - Qiniu Cloud's platform has over 169,000 developers, with a rapid increase in AI-related user demand, reaching 15,000 users due to the availability of over 50 adjustable large models [3][4] - Future collaborations will explore vertical fields such as AI + education and AI + energy, aiming to create new opportunities in the AI sector [4]

Core Viewpoint - Qiniu Intelligent (2567.HK) has entered into a strategic partnership with Wuxiang Cloud Valley to scale AI inference computing power, targeting the trillion-dollar AI inference computing market [1] Group 1: Strategic Partnership - The collaboration aims to provide accessible AI inference computing power for enterprises and entrepreneurs [1] - Both companies will explore vertical fields such as AI + education and AI + energy [1] Group 2: Product Offerings - Qiniu Intelligent currently has three product lines: Media Cloud, AI Cloud, and LinX Cloud, with a new focus on AI inference computing products [1] - The launch of LinX AI in August signifies Qiniu's comprehensive layout from cloud to edge, expanding into the smart hardware AI market [1] Group 3: Market Demand - As of August, Qiniu Intelligent's developer community has exceeded 1.69 million, with a continuous increase in new registrations [1] - The demand for AI computing power is rising, and the partnership will enable Qiniu to offer refined large model AI inference and computing power leasing services [1] - This collaboration is expected to lower the development threshold for AI applications, significantly shortening the cycle from concept validation to large-scale commercialization [1]

Core Insights - Qiniu Cloud (02567) has entered a strategic partnership with Wuxiang Cloud Valley to scale AI inference computing power, targeting the trillion-dollar AI inference market [1][3] - The collaboration aims to provide affordable AI inference computing power for enterprises and entrepreneurs, enhancing the accessibility of AI technologies [1][4] Company Overview - Qiniu Cloud has 14 years of experience in the audio and video sector, with three main product lines: Media Cloud, AI Cloud, and LinX Cloud, marking a comprehensive layout from cloud to edge [1][3] - In the first half of 2025, Qiniu Cloud's AI-related revenue reached 184 million yuan, accounting for 22.2% of total revenue, primarily from AI inference services and computing resource leasing [3] Market Dynamics - The demand for AI inference is rapidly increasing, driven by over 50 deployable large models on Qiniu Cloud's platform, including LLM inference models and AI programming support [3] - The number of AI-related users on Qiniu Cloud has surged to 15,000, indicating a strong market interest and growth potential in AI inference services [3] Infrastructure and Investment - Wuxiang Cloud Valley's Intelligent Computing Center, with a total investment of 3.6 billion yuan and a construction area of 149,000 square meters, aims to provide up to 40,000P of intelligent computing power [4] - The partnership is expected to unlock the commercial potential of Wuxiang Cloud Valley's computing resources, enhancing the overall service offerings in AI [4] Future Prospects - Qiniu Cloud and Wuxiang Cloud Valley plan to explore collaborations in vertical fields such as AI + education and AI + energy, aiming to create new opportunities in the AI landscape [4]

Summary of Conference Call Records Industry or Company Involved - Focus on the semiconductor industry, brain-computer interfaces, and deep-sea technology [1][16] Core Points and Arguments - The A-share market is under pressure due to multiple factors including the escalation of the Middle East conflict, policy expectations from the Lujiazui Forum, and technical resistance levels in the Shanghai Composite Index [1][2] - The macroeconomic recovery is weak, with May retail sales data exceeding expectations but lacking sustainability; the equipment manufacturing sector is performing relatively well, while real estate and other sectors face ongoing pressure [1][2] - Structural issues within the market are evident, with small-cap stocks showing excessive returns, leading to increased crowding and a decline in large-cap performance, raising market fragility [1][2] - Funds are shifting from TMT (Technology, Media, and Telecommunications) in Q1 to new consumption and innovative pharmaceuticals in Q2, impacting market performance and trading volume [1][3] - The Hong Kong stock market is experiencing significant liquidity-driven characteristics, with concerns over the Hong Kong Monetary Authority tightening liquidity due to the depreciation of the Hong Kong dollar [1][3] - The semiconductor industry is performing strongly, supported by policies that facilitate the listing of unprofitable hard-tech companies on the STAR Market, and increasing expectations for domestic production in the semiconductor materials and equipment sectors due to U.S.-China tensions [1][12] Other Important but Possibly Overlooked Content - The A-share market is showing weakness, with only dividend or blue-chip value sectors performing relatively well; funds are flowing into defensive sectors like banks and insurance due to increased risk aversion [1][5] - The recent policy change shortening the clinical trial application period for innovative drugs to 30 days exceeded most investors' expectations, but the market's response has been muted due to the current phase of the market [6][7] - The market sentiment index has declined significantly, indicating a drop in active investor confidence, with fewer stocks hitting the daily limit up and a stable number of limit down stocks [8] - The computing power communication industry has gained attention due to the rebound in early June, with a focus on companies with high earnings visibility as the market approaches mid-year reports [10] - The semiconductor sector is seeing strong performance, particularly in manufacturing equipment and storage, with expectations for further listings of quality companies in the semiconductor supply chain [12] - The renewable energy sector is also showing signs of recovery, particularly in solar energy and solid-state batteries, with increased interest from investors [13] - The liquor industry has rebounded after a period of pressure, aided by media corrections that alleviated overly pessimistic market sentiment [14] - Developments in the stablecoin sector are noteworthy, with major companies entering the market, although the direct impact on A-share fundamentals remains limited [15] - Investment recommendations include focusing on the semiconductor sector, core companies in computing power communication, AI applications, and emerging technologies like deep-sea technology and brain-computer interfaces [16][18]