1月26日，市场研究机构Counterpoint Research在最新研报中预测，非GPU服务器AI芯片——AI ASIC阵 营在近期将经历高速增长，到2027年出货数量将达到2024年的三倍，2028年则有望以1500余万颗的规模 反超GPU。 报告显示，这一爆炸性增长的背后，是对谷歌TPU基础设施的强劲需求、AWS Trainium集群的持续扩 展，以及Meta(MTIA)和微软(Maia)随着其内部芯片产品组合的扩展而带来的产能提升。 就人工智能超大规模数据中心的出货量而言，谷歌预计将在2027年之前保持市场领先地位，这主要得益 于其Gemini生态系统的爆炸式增长。 针对谷歌的统治地位，Counterpoint Research研究员David Wu强调："尽管由于市场规模不断扩大，以及 竞争对手(如博通、Marvell和Alchip等设计公司)纷纷采用自研芯片，预计谷歌的市场份额将在2027年下 降至52%，但其TPU集群仍将是无可争议的行业核心和标杆。这一基准的支撑源于训练和运行下一代 Gemini模型所需的庞大且持续的计算能力，而这又需要谷歌持续、积极地扩充其内部芯片基础设施。" 该机构认为， ...

Core Insights - North American cloud service providers (CSPs) are significantly increasing investments in AI infrastructure, leading to a projected annual growth rate of over 28% in global AI server shipments by 2026 [1] - The demand for AI inference services is driving a replacement and expansion cycle for general servers, with an expected annual growth rate of 12.8% in global server shipments (including AI servers) by 2026 [1] Group 1: Market Trends - From 2024 to 2025, the server market will focus on training advanced large language models (LLMs) using AI servers equipped with GPUs and HBM for parallel computing [3] - Starting in the second half of 2025, the development of AI inference services such as AI Agents, LLaMA model applications, and Copilot upgrades will prompt CSPs to shift towards monetization and profit models [3] - The total capital expenditure growth rate for the five major North American CSPs (Google, AWS, Meta, Microsoft, Oracle) is projected to reach 40% in 2026, driven by large-scale infrastructure investments and the replacement of general servers purchased during the 2019-2021 cloud investment boom [3] Group 2: AI Server Market Dynamics - The 2026 AI server market will be primarily driven by North American CSPs, government sovereign cloud projects, and large CSPs accelerating their own ASIC development and edge AI inference solutions [4] - GPUs are expected to account for 69.7% of AI chip usage, with NVIDIA's GB300 models becoming the mainstream for shipments, while VR200 will gradually ramp up in the second half of the year [4] Group 3: ASIC Development - The share of ASIC AI servers in shipments is expected to rise to 27.8% by 2026, the highest since 2023, with growth rates surpassing those of GPU AI servers [6] - Google is leading the investment in self-developed ASICs, with its TPU not only serving Google Cloud Platform infrastructure but also being sold to external companies like Anthropic [6]

Core Insights - Bitdeer Technologies Group reported an increase in self-mining hashrate to 41.2 EH/s, surpassing its target of 40 EH/s, driven by the deployment of SEALMINER mining rigs [1][4][6] - The company mined 511 Bitcoins in October 2025, reflecting a 13% increase from September 2025 [4][7] - Bitdeer achieved an annual recurring revenue (ARR) of US$8 million from its AI cloud services, supported by strong customer demand for NVIDIA B200 systems [5][6] Mining Operations - The total proprietary hash rate deployed reached 41.3 EH/s in October 2025, up from 35.0 EH/s in September 2025 [2][7] - The company has 254,000 mining rigs under management, with 166,000 self-owned and 88,000 hosted [7] - The total hash rate under management increased to 55.5 EH/s, compared to 49.2 EH/s in September 2025 [7] SEALMINER Development - The SEALMINER A3 and A2 models are in final assembly, with the A3 model achieving a hashrate of 0.3 EH/s and the A2 model at 2.6 EH/s [2] - The first SEAL04 chip demonstrated power efficiency of approximately 6-7 J/TH, with mass production targeted for Q1 2026 [5][6] Infrastructure Updates - The company has completed construction of several data centers, including a 175 MW site in Tydal, Norway, and a 50 MW site in Oromia, Ethiopia, with 40 MW already energized [10][15] - Ongoing projects include a 221 MW site in Massillon, Ohio, expected to be fully energized by Q1 2026 [13][15] - The total global electrical capacity across all sites is 2,992 MW, with additional pipeline capacity of 1,381 MW [12][14] AI Cloud Services - Bitdeer deployed 584 GPUs with an 87% utilization rate, indicating strong demand for its AI cloud services [5][6] - The company is expanding its GPU infrastructure and has placed orders for NVIDIA's next-generation systems, expected to be delivered in December 2025 [5][6]

Financial Data and Key Metrics Changes - Q1 fiscal year 2026 revenue was $5 billion, down 15% year-over-year and down 13% quarter-over-quarter, compared to guidance of $6 billion-$7 billion [15] - Non-GAAP gross margin for Q1 was 9.5% versus 9.6% in Q4 [17] - GAAP diluted EPS was $0.26 compared to guidance of $0.30-$0.42, and non-GAAP diluted EPS was $0.35 versus guidance of $0.40-$0.52 [18] - Cash flow used in operations for Q1 was $918 million compared to cash flow generated from operations of $864 million in the prior quarter [18] - Q1 closing inventory was $5.7 billion, up from $4.7 billion in Q4 [18] Business Line Data and Key Metrics Changes - AI GPU platforms represented over 75% of Q1 revenues, continuing to be the key growth driver [15] - Enterprise channel revenues totaled $1.5 billion, representing 31% of revenues, down from 36% in the prior quarter [15] - OEM appliance and large data center segment revenues were $3.4 billion, representing 68% of Q1 revenues, up from 63% in the last quarter [16] Market Data and Key Metrics Changes - By geography, the U.S. represented 37% of Q1 revenues, Asia 46%, Europe 14%, and the rest of the world 3% [16] - Year-over-year, U.S. revenues decreased 57%, while Asia grew 143% [16] - Asia's growth was attributed to a U.S.-based customer opening a large data center in Asia [16] Company Strategy and Development Direction - The company is focused on leading the AI growth trend with innovative solutions, including the DCBBS [7] - The DCBBS is critical for future success, enabling rapid planning, design, and deployment of AI-ready data centers [9] - The company is executing an aggressive global expansion, with new production facilities in Taiwan, the Netherlands, Malaysia, and soon the Middle East [11] - The long-term goal is to expand revenues in higher margin segments such as data center building block solutions and software service offerings [21] Management's Comments on Operating Environment and Future Outlook - Management expects to ship at least $10.5 billion in the December quarter, with a full-year revenue outlook raised to at least $36 billion [14][22] - The company anticipates a sequential growth through fiscal 2026, driven by strong demand for AI infrastructure [14] - Management acknowledged the challenges of ramping production and the complexities of new GPU racks impacting revenue timing [8][15] Other Important Information - The company executed a $1.8 billion AR facility to strengthen working capital [19] - The Q1 cash conversion cycle was 123 days, up from 96 days in Q4 [19] - Capital expenditures for Q1 totaled $32 million, resulting in negative free cash flow of $950 million for the quarter [18] Q&A Session Summary Question: What is driving the increased revenue guidance? - Management indicated that the availability of NVIDIA Blackwell Ultra chipsets is a key driver for increased revenue guidance, alongside market share expansion [26] Question: When will DCBBS impact gross margins? - DCBBS has received positive feedback, and management expects it to ramp up quickly, contributing to higher profit margins in the future [28] Question: What components contribute to the strong order outlook? - The strong order outlook is driven by GPU products like Blackwell Ultra and AMD MI350, along with a focus on providing end-to-end data center solutions [32] Question: Is there conservatism in the revenue guidance? - Management confirmed that the revenue guidance is conservative, with expectations for continued growth as capacity and technology mature [42] Question: How is the company managing working capital? - The company has implemented an accounts receivable sale program to enhance working capital flexibility, especially as revenues double [56] Question: What is the status of the Malaysia plant? - The Malaysia plant is starting to ramp up production and is expected to contribute significantly going forward [51] Question: How does the company plan to improve margins? - Management is focused on leveraging additional business, expanding manufacturing in various geographies, and enhancing DCBBS strategies to improve margins [75]

Summary of Key Points from the Conference Call on Liquid Cooling Market Industry Overview - The discussion focused on the **liquid cooling industry**, particularly in the context of **AI technology** and its applications in data centers and computing power management [1][3]. Core Insights - **Full Liquid Cooling Adoption**: The trend towards full liquid cooling systems is increasing, with NVIDIA's GB300 model fully adopting this technology, while previous models like GB200 still partially relied on air cooling [3]. - **Cloud Service Providers (CSPs)**: Major CSPs such as Google, Amazon AWS, and Meta are considering full liquid cooling solutions due to the rising computational demands of ASICs. Full liquid cooling systems are noted for their long lifecycles in AI data centers [3]. - **Technological Innovations**: - NVIDIA's Microchannel Liquid Cooling Plate (MLCP) and Microsoft's in-chip Microfluidic cooling system are highlighted as significant advancements in cooling technology [1][2]. - MLCP utilizes micro channels for liquid coolant flow, directly attached to computing chips, presenting high entry barriers due to the complexity of smaller channel sizes [3]. - The Microfluidic solution is more advanced, requiring IC-level production processes, which necessitates collaboration with leading foundries [3]. - **Market Dynamics**: The expert indicated that MLCP may see earlier adoption compared to the Microfluidic solution due to existing technological challenges [2][3]. Additional Important Points - **Installation and Maintenance**: The modularized Coolant Distribution Unit (CDU) is easier to install and maintain, which is a significant advantage for CSPs [3]. - **Competitive Landscape**: The competitive dynamics in the liquid cooling market are evolving as more companies adopt these advanced cooling solutions to meet the demands of high-performance computing [1][3]. This summary encapsulates the key takeaways from the conference call regarding the liquid cooling market and its implications for the technology and AI sectors.

Core Viewpoint - IREN Limited has achieved NVIDIA Preferred Partner status and expanded its GPU fleet with the acquisition of additional NVIDIA GPUs, enhancing its AI Cloud capabilities [1][3]. Group 1: GPU Acquisition and Financing - IREN has procured 1.2k air-cooled NVIDIA B300s and 1.2k liquid-cooled NVIDIA GB300s for approximately $168 million, increasing its total GPU fleet to 10.9k NVIDIA GPUs [1][3]. - The financing for the GB300s amounts to approximately $96 million, structured as a 24-month lease covering 100% of the purchase price, with lease payments based on a high single-digit interest rate [2]. - The B300 GPUs are funded from existing cash, with additional financing efforts in progress [2]. Group 2: Data Center Expansion - The new GPUs will be installed at IREN's Prince George campus, where a 10MW liquid-cooled installation is under construction, capable of supporting over 4.5k GB300s [3]. - The data center platform supports both liquid-cooled and air-cooled systems, backed by nearly 3GW of grid-connected power, scalable to hundreds of thousands of GPUs [4]. Group 3: Company Overview - IREN is a leading developer and operator of next-generation data centers, focusing on Bitcoin, AI, and renewable energy solutions [8]. - The company operates 810MW of data centers across three verticals: Bitcoin Mining, AI Cloud Services, and AI Data Centers, with a significant focus on high-performance computing [10].

Group 1 - TrendForce indicates that the demand for MLCC orders may be impacted due to the gradual decline of preemptive consumption and inventory accumulation, with the back-to-school season in Q3 facing uncertainties [1] - ODM orders for mid-to-low-end consumer products such as smartphones, laptops, and tablets are expected to remain flat or increase by only about 5% in Q3, reflecting a conservative order strategy [1] - AI Server orders are experiencing a surge, driven by the simultaneous release of NVIDIA's GB200 and GB300 platforms, leading to a nearly 25% average quarterly increase in MLCC stocking demand [1] Group 2 - The capacity utilization rates among MLCC suppliers show significant disparities, with Japanese and Korean manufacturers focusing on high-end AI applications averaging 90%, while Chinese manufacturers are around 75% [2] - MLCC suppliers are accelerating the establishment of testing and packaging production lines in Southeast Asia to achieve localized production [2] - OEMs are expected to face cost pressures, leading to potential price increases for end products as they release RFQ for 2026 mobile phones and laptops [2]

Core Insights - The article highlights the impact of early consumption and inventory accumulation on the demand for MLCCs, indicating potential fluctuations in back-to-school season consumption in Q3 2023 [1] - There is a noticeable polarization in industry demand, with ODM orders for mid-to-low-end consumer products showing only a slight increase, reflecting a conservative approach to order placements [1] - The demand for AI servers is surging, driven by the simultaneous release of NVIDIA's GB200 and GB300 platforms, positively affecting the revenue of major ODMs [1] Group 1 - TrendForce reports that MLCC order demand may be affected by the diminishing early consumption and inventory accumulation, leading to uncertainties in the Q3 back-to-school season [1] - ODM orders for mid-to-low-end consumer products are expected to remain flat or increase by only about 5% in Q3, indicating a shift towards a more cautious order strategy [1] - Many companies have preemptively shipped products in the first half of the year to adapt to international market changes, which has depleted traditional demand for the second half of the year [1] Group 2 - The supply chain's operational rates are polarized, with Japanese and Korean manufacturers focusing on high-end AI applications achieving an average capacity utilization rate of 90%, while Chinese manufacturers are around 75% [2] - MLCC suppliers are accelerating the establishment of testing and packaging lines in Southeast Asia to localize production and supply [2] - OEMs are expected to face cost pressures, leading to potential price increases for end products as they release RFQ for 2026 mobile phones and laptops [2]

Core Insights - Micron Technology's high bandwidth memory (HBM) chips are gaining significant traction due to their increasing application in high-performance computing, hyperscalers, and AI data centers, with HBM revenues surpassing $1 billion in Q2 FY25 [2][10] - The company is fully sold out of its HBM supply for calendar 2025, driven by strong demand for its HBM3E and HBM3E-12H variants, which offer better power efficiency and higher memory capacity compared to competitors [3][10] - Micron's HBM market is projected to exceed $35 billion in 2025, with a focus on securing delivery agreements for 2026 [4] Company Performance - Micron is set to ramp up production of the next generation HBM4 in 2026, which will provide over 60% more bandwidth than HBM3E [5] - The company anticipates Q3 FY25 revenues of $8.80 billion, reflecting a year-over-year growth of 29.33% [6] - Micron's shares have increased by 31.8% year-to-date, outperforming the Zacks Computer - Integrated Systems industry, which grew by 15.1% [9] Competitive Landscape - The global HBM market is highly consolidated, with key players including SK hynix, TSMC, and Samsung, although not all directly compete with Micron [7] - SK hynix and TSMC are collaborating to develop HBM4 and next-generation packaging technology, which presents growth opportunities for Micron [8] Valuation and Earnings Estimates - Micron trades at a forward price-to-sales ratio of 2.84X, lower than the industry average of 3.53X [11] - The Zacks Consensus Estimate for Micron's fiscal 2025 earnings indicates a year-over-year growth of 433%, with a 55.87% growth estimate for 2026 [12]