Company and Industry Summary Company: Ta Liang Key Points - **Upgrade of Communication Protocol**: The upcoming NVIDIA Rubin platform will feature an upgraded communication protocol between GPU and NVSwitch, utilizing 224GSerDes on PCB with a significant signal frequency increase to a maximum of 56GHz, compared to 28GHz on the Blackwell platform. This upgrade indicates that even minor via stubs can lead to severe signal degradation, establishing a ±2mil stub height as the threshold for signal transmission quality. Accurate closed-loop measurement is deemed crucial for manufacturing quality control [1][1][1] - **StubMapper Device**: Ta Liang has developed a proprietary device called StubMapper, which enables ±2mil depth measurement to guide subsequent back-drilling processes executed by Ta Liang's 6-axis CCD drilling machine. StubMapper can provide profile measurement data back to the back-drilling machine, enhancing precision in manufacturing [1][1][1] - **Increased Demand for Back-Drilling**: There has been a surge in industry demand for back-drilling to meet the requirements of high-end AI PCBs, resulting in an extension of PCB equipment delivery cycles from 1.5 months to over 3 months. This demand has led Ta Liang to consider outsourcing more low-end drilling tool production to focus on high-end tool manufacturing [2][2][2] - **Production Capacity and Future Plans**: Ta Liang's current monthly production capacity is approximately 300 sets of drilling tools. The company plans to enhance the throughput of StubMapper by introducing a four-probe version, with mass production scheduled for March 2026, and initial customer reservations already in place [2][2][2] - **Semiconductor Equipment Division**: Ta Liang also operates a semiconductor equipment division focused on measurement and automated optical inspection (AOI), offering 7 types of inspection equipment used in processes such as HBM incoming material and flip-chip bonding. This division has received certification from a leading foundry for CoWoS and SoIC platforms, with CoWoS-related revenue accounting for approximately 75% of its semiconductor equipment business [2][2][2] - **Milestone Achievement**: The certification from the leading foundry is considered a key milestone for Ta Liang, facilitating smoother entry into the OSAT supply chain as OSAT manufacturers expand their 2.5D capacity [2][2][2]

Core Insights - In early 2025, significant developments in the AI chip sector were reported, including Elon Musk's confirmation of Tesla's (TSLA.US) revival of the Dojo 3 supercomputer project, aiming to become the largest AI chip manufacturer globally, and Cerebras Systems' multi-year procurement agreement with OpenAI worth over $10 billion, promising 750 megawatts of computing power by 2028 [1][2]. Group 1: AI Chip Evolution - The evolution of AI chips is characterized by two distinct designs: Cerebras' wafer-scale integration and Tesla's Dojo, which represents a hybrid approach between single-chip and GPU clusters [3]. - The divergence stems from different solutions to the "memory wall" and "interconnect bottleneck" challenges, with traditional GPU architectures facing limitations in memory bandwidth compared to computational power [3][4]. Group 2: Cerebras' Innovations - Cerebras' WSE-3 chip features 40 trillion transistors, 900,000 AI cores, and 44GB of on-chip SRAM, achieving a bandwidth of 214 Pb/s, significantly outperforming NVIDIA's H100 [4]. - The design addresses yield issues associated with large wafers by minimizing the size of each AI core and employing redundancy to maintain performance despite defects [4]. Group 3: Tesla's Strategic Shift - Tesla's Dojo project faced setbacks but was revived with a new focus on "space AI computing," moving away from its original goal of competing with NVIDIA's GPU clusters [7][8]. - The AI5 chip, designed with a 3nm process, is expected to be produced by the end of 2026, aiming for performance comparable to NVIDIA's Hopper architecture [8]. Group 4: Market Dynamics and Competition - The AI chip market is becoming increasingly crowded, with competitors like AMD and NVIDIA rapidly advancing their offerings, which poses challenges for alternative architectures like wafer-scale systems [16][19]. - Cerebras aims to differentiate itself by focusing on low-latency inference systems, capitalizing on the growing demand for real-time AI applications [16][14]. Group 5: Strategic Partnerships - Cerebras' partnership with OpenAI, involving a $10 billion commitment for computing power, highlights the increasing importance of low-latency inference capabilities in the AI landscape [11][12]. - The collaboration reflects a broader trend of established tech companies integrating promising AI chip startups into their ecosystems, which may reshape the competitive landscape [20][21].

Group 1 - The launch of NVIDIA's Rubin platform marks a new era in AI computing, significantly enhancing inference performance and reducing costs through the collaboration of six modules, including Vera CPU and Rubin GPU [1] - The Rubin GPU features a Transformer engine, achieving NVFP4 inference performance of 50 PFLOPS, which is five times that of Blackwell; the BlueField-4 powered memory storage platform expands GPU memory capacity and supports high-speed sharing, increasing tokens per second by up to five times [1] - The PCB segment adopts a cable-free interconnection architecture, improving assembly speed by 18 times and lowering deployment and operational costs; the open-source model ecosystem covers six major fields, including biomedicine and physical simulation [1] Group 2 - The Guotai ETF (159388) tracks the ChiNext AI Index (970070), which has a daily price fluctuation limit of 20%; this index selects listed companies involved in software development, smart hardware, and data services within the AI technology and application sectors to reflect the overall performance of AI-related stocks in the ChiNext market [2] - The index constituents exhibit high growth and innovation characteristics, focusing on the forefront of China's AI industry development [2]

Key Insights - The report highlights the launch of NVIDIA's Rubin platform, which significantly reduces the cost of generating tokens to about one-tenth of the previous generation, Blackwell, and is now in full production. This development is expected to drive demand in the semiconductor industry, with TSMC reporting a record revenue growth of 31.6% year-on-year for 2025, indicating a robust recovery in semiconductor demand [5][6][7]. Group 1: Electronic Industry Insights - NVIDIA's Rubin platform was showcased at CES 2026, featuring six chip components that enhance AI deployment efficiency, with the platform now fully in production [5][6]. - TSMC reported a December revenue of approximately NT$335 billion, a year-on-year increase of 20.4%, contributing to a total annual revenue of NT$3.81 trillion, marking a 31.6% increase and setting a historical high [7][8]. - The electronic sector is experiencing a demand recovery, with rising prices for memory chips and a strong push for domestic production in China, suggesting structural investment opportunities in AI computing, AIOT, semiconductor equipment, and key components [5][9]. Group 2: Machinery Equipment Industry Insights - The report indicates a significant increase in excavator sales, with December 2025 sales reaching 23,095 units, a year-on-year growth of 19.2%, driven by both domestic and export markets [11][12]. - The total excavator sales for 2025 were 235,257 units, reflecting a 17% increase, with domestic sales up by 17.9% and exports up by 16.1% [13][14]. - Loader sales also saw substantial growth, with December 2025 sales of 12,236 units, a 30% increase year-on-year, and total sales for the year reaching 128,067 units, up 18.4% [12][14]. Group 3: U.S. Economic Recovery Insights - The U.S. labor market showed signs of recovery, with December 2025 non-farm payrolls increasing by 50,000, although slightly below expectations, while the unemployment rate fell to 4.4% [17][18]. - The service sector contributed significantly to job growth, particularly in leisure and hospitality, which rebounded due to the holiday season [19][20]. - The report suggests that the U.S. economy is on a recovery path, with expectations for continued job growth and a potential decrease in interest rates by the Federal Reserve [21][22].

Summary of Key Points from the Conference Call Company and Industry Overview - The conference call discusses the advancements in liquid cooling technology, specifically focusing on NVIDIA's Rubin platform and its components, including cold plates and related technologies [1][2][3]. Core Insights and Arguments - **Rubin Platform Design**: The Rubin platform utilizes a large plate model combined with microchannel technology to enhance GPU cooling efficiency, achieving 100% liquid cooling coverage for compute and switch trays [1][3]. - **Cold Plate Value Increase**: The introduction of microchannel technology and galvanized coating techniques has increased the value of cold plates for each compute tray to an estimated $500-$550, with expectations to drop below $500 by 2027 [1][5]. - **Microchannel Technology**: This technology, which embeds micron-sized water flow channels into cold plates via laser welding, significantly improves cooling efficiency but comes with high costs and limited current adoption [1][6][7]. - **Market Share Dynamics**: Taiwanese suppliers continue to dominate the cold plate market, but domestic suppliers like Invec and Siquan are expected to gain market share, potentially reaching 20% in the Ruby series [1][11]. - **Transition to Bidirectional Cooling**: The industry trend is shifting from unidirectional to bidirectional cold plates to enhance heat dissipation and conduction efficiency, although this technology faces challenges in precise control of liquid circulation systems [1][13][15]. Additional Important Content - **Supplier Selection Criteria**: NVIDIA prioritizes brand reputation, supply reliability, and pricing when selecting cold plate suppliers, favoring well-known brands unless suitable alternatives are unavailable [1][10]. - **Component Value Breakdown**: In a $70,000 72-node cabinet, cold plates account for approximately $32,000 to $34,500, while manifold components contribute around $25,000 [1][21]. - **Impact of New Suppliers**: The introduction of new suppliers has minimal impact on profit margins due to transparent pricing and direct negotiations between clients and suppliers [1][9]. - **Future Design Changes**: The Ruby Ultra may require design modifications to accommodate increased GPU power demands, potentially leading to more advanced cooling solutions [1][12]. - **Control of Key Components**: NVIDIA retains decision-making authority over the procurement of critical components like cold plates, ensuring assembly quality and consistency [1][22]. This summary encapsulates the essential points discussed in the conference call, highlighting the advancements in liquid cooling technology and the strategic positioning of NVIDIA and its suppliers in the market.

Group 1 - The core viewpoint of the report is that 2026 will witness strong growth in the AI supply chain, driven by the expansion of capital expenditures from hyperscale cloud service providers, rapid advancements in AI hardware technology, and a tight supply chain for components [1] Group 2 - Capital expenditures from top global cloud service providers are expected to increase by 34% year-on-year in 2026, reaching approximately $424 billion, primarily fueled by the ongoing expansion of cloud services and internet businesses, with cloud service revenue growth rebounding to over 28% [2] Group 3 - The report predicts a prosperous coexistence of AI GPUs and custom chips (ASICs) in 2026, with Nvidia's Blackwell platform expected to enter large-scale delivery, anticipating the delivery of around 60,000 racks throughout the year [3] - Google and Amazon are leading the deployment of custom chips, with Google's TPU v7 and Amazon's Trainium T3 expected to be widely deployed in 2026, while Meta plans to launch its ASIC solution in the second half of the same year [3] Group 4 - Traditional general-purpose servers are also showing a recovery growth in the high single to mid-single digits, alongside the AI boom [4] Group 5 - The report highlights an uneven distribution of growth benefits across the supply chain, with markets for rack power supplies, cooling, and PCB substrates expected to continue expanding, while tight supply of various raw materials, especially memory, creates a "seller's market" benefiting component suppliers [5] - Due to high prices of storage components, the report has downgraded the 2026 PC shipment forecast, expecting a 4% decline, while smartphone growth is also expected to slow to 2% due to rising commodity prices [5] Group 6 - The report suggests that the driving factors for stock prices in 2026 will be sales and profit growth rather than further valuation expansion, identifying the best opportunities among suppliers that can benefit from the increase in AI server deployments and enhance unit value or added value [6] - Recommended stocks include server ODM manufacturers (Quanta, Hon Hai, Wistron, Wistron NeWeb) and component suppliers (Delta, Jentech, AVC), as well as Largan Precision, which benefits from upgrades in foldable screens and variable aperture specifications [6] - The report takes a more cautious stance on brand companies (ASUS, Lenovo, Gigabyte, MSI) and ODM and component companies (Compal, Pegatron, Inventec, JMicron) that are less likely to benefit from AI server growth due to high input costs, especially in storage [6]

Macro - The People's Bank of China plans to implement a moderately loose monetary policy by 2026, establishing a mechanism to provide liquidity to non-bank institutions under specific scenarios to support capital markets [1] Satellite Internet - The first commercial rocket company in China, Kegong Rocket, is set to change ownership with an overall valuation of approximately 11.2 billion yuan [2] Semiconductor - NVIDIA's BlueField-4 is expected to reconstruct the AI storage hierarchy, potentially opening up new demand for NAND [3] - NVIDIA's CEO Jensen Huang announced at CES that the era of "physical AI" has arrived [7] Pricing Strategies - Seagate plans to increase product prices by 10% each quarter [4] Non-ferrous Metals - China has announced a ban on all dual-use items for military purposes to Japan, affecting critical metal materials such as rare earths, tungsten, molybdenum, indium, lithium, and titanium alloys [5] Artificial Intelligence - Market discussions regarding NVIDIA's next-generation Rubin platform specifications indicate no issues with downgrading [6] Robotics - Tesla's official recruitment platform has posted job openings related to "covering materials and soft products" for the Optimus robot [8] New Energy - A meeting for the power and energy storage battery industry is scheduled for January 7 [9] - Feedback from the industry indicates extremely low inventory levels of lithium carbonate and strong demand, with two listed companies in lithium iron phosphate confirming price increases [10] - Australia has significantly increased subsidies for home battery systems [11] - The organic silicon industry plans to intensify production cuts [12] Brain-Computer Interface - The brain-computer interface "unicorn" Qiangna Technology has completed a financing round of 2 billion yuan [13] Intelligent Driving - NVIDIA has released an open-source model family for autonomous driving, aimed at addressing long-tail issues and accelerating the implementation of intelligent driving [14] Market Dynamics - Guosheng Technology and Jiamei Packaging are under suspension for verification; Fenglong Co. has issued a warning about stock fluctuations; several commercial aerospace and brain-computer interface companies have released announcements regarding stock movements [15] - In December 2025, the number of new A-share accounts opened on the Shanghai Stock Exchange increased by approximately 9% month-on-month and surged by 30.55% year-on-year [15] Strategy Observation - On Tuesday, trading volume reached 2.81 trillion yuan, with a significant increase in market sentiment; however, rapid gains and the activation of financial sectors may raise concerns about the "healthy bull" rhythm [16] - The regulatory authorities have released signals of concern through suspensions and announcements, which may provide divergence opportunities for the market, but substantial cooling requires sustained measures [16] - Sectors such as non-ferrous metals, non-bank financials, and chemicals led the gains, indicating active institutional investment [16] - The thematic sectors, particularly those related to Musk's industries (intelligent driving and robotics), performed actively; while the semiconductor sector (e.g., storage) has solid logic, it has not yet become a focal point, with potential for future attention [16] - Currently, large funds tend to operate around mainline capacity varieties [16]

Core View - Coreweave (CRWV.US) shares rose over 2% pre-market, reaching $78.44, following the announcement from NVIDIA CEO Jensen Huang at CES regarding the full-scale production of the new flagship AI computing platform, Vera Rubin [1] Company Developments - CoreWeave announced the integration of NVIDIA's Rubin technology into its AI cloud platform, aimed at expanding options for clients developing intelligent agent AI, inference, and large-scale inference workloads [1] - The company expects to be one of the first cloud service providers to deploy the NVIDIA Rubin platform in the second half of 2026 [1] Analyst Ratings - D.A. Davidson upgraded CoreWeave's rating from "underperform" to "neutral," significantly raising the target price from $36 to $68, indicating that the stock price may benefit from the substantial financing actions of its client, OpenAI [1]

Group 1: Technology Developments - NVIDIA launched a new generation of AI platform Rubin, featuring six new chips, including Rubin GPU with a reasoning power of 50 PFLOPS, five times that of Blackwell [2] - Qualcomm and Google announced an expansion of their decade-long partnership to create end-to-end automotive technology solutions, integrating Qualcomm's Snapdragon digital chassis with Google's automotive software [4] - NVIDIA introduced the Alpamayo series of open-source AI models and tools aimed at enhancing the development of safe and reliable autonomous driving systems [11] Group 2: Corporate Collaborations - Zhiyuan Robotics partnered with MiniMax to enhance voice interaction experiences through customized AI technology [5] - GAC Group signed a comprehensive cooperation framework agreement with Huawei to deepen collaboration on smart vehicle technologies [5] Group 3: Market Movements - Chinese aerospace company China Aerospace Science and Technology Corporation increased its registered capital from 1 billion to 1.396 billion yuan, a 77.6% increase [10] - Defu Technology announced a cooperation agreement with a leading domestic CCL company to supply high-end copper foil products, expected to positively impact its 2026 performance [9] Group 4: Product Launches and Innovations - Raybird Innovation completed over 1 billion yuan in financing and plans to showcase the world's first AR smart glasses with eSIM technology at CES 2026 [15] - ByteDance's Doubao denied rumors regarding the imminent shipment of AI glasses, clarifying that there are currently no sales plans [8]

Group 1 - A passenger lost a DJI Pocket 3 camera worth over 3,000 yuan in a ride with Cao Cao Chuxing, and the driver refused to return it after receiving a payment of 50 yuan [2] - Cao Cao Chuxing has permanently banned the driver's account and is cooperating with the passenger to report the incident to the police while promising to compensate the passenger [2] Group 2 - Xiaomi's CEO Lei Jun responded to claims about the "wheel loss protection" design in their cars, stating it is a mature safety design used in many luxury vehicles [2][3] - The "wheel loss protection" strategy aims to prevent the wheel hub from causing secondary injuries to passengers during severe collisions [3] Group 3 - Venezuela's oil exports have nearly dropped to zero due to U.S. sanctions, leading to a saturation of domestic oil storage facilities [4][5] - The Venezuelan state oil company has begun to cut crude oil production and is implementing emergency measures to close some oil fields [4] Group 4 - The 2026 childcare subsidy application started on January 5, allowing eligible applicants to apply online or offline [6] - The new application process includes a renewal feature for those who applied in 2025 and meet the criteria for the new year [6] Group 5 - Microsoft China’s chairman Hou Yang denied rumors of leaving for ByteDance, clarifying that he is not considering a job change [6] - Tencent's AI product, Yuanbao, faced backlash for generating offensive responses, and the company is investigating the issue [6] Group 6 - The founder of Double Star Mingren, Wang Hai, announced a severance of ties with his son and daughter-in-law due to their U.S. citizenship [7] - The company, established in 1921, is one of China's earliest national shoe manufacturing brands [7] Group 7 - A court ruling ordered Xu Xiang and Xu Changjiang to compensate 23 investors over 3.29 million yuan, with the company Wenfeng Shares bearing joint liability [8] Group 8 - The company Pangu Technology announced the completion of over 100 million yuan in Pre-A round financing to support RRAM technology research and team expansion [19] - The AI automation platform Jiukexin completed B2 round financing, raising over 100 million yuan for product development and market expansion [20] Group 9 - Tencent's AI workbench, ima, launched a new feature for generating PowerPoint presentations, allowing users to create slides using personal knowledge bases [24] - NetEase Youdao's AI answering pen introduced a generative video explanation feature, marking a significant upgrade in AI learning hardware [25]