Core Viewpoint - The HBM (High Bandwidth Memory) market is expected to experience tight supply and demand until 2027, driven by technological iterations and AI demand, with SK Hynix and Micron leading the market due to their technological and capacity advantages [1][2]. Supply and Demand Trends - HBM supply tightness is projected to persist through 2027, with a gradual easing of oversupply expected in 2026-2027. Channel inventory is anticipated to increase by 1-2 weeks, reaching a healthy level [2]. - The delay in Samsung's HBM certification and the strong demand growth from NVIDIA's Rubin GPU are the main factors contributing to the current supply-demand tension [2]. - HBM4 supply is expected to significantly increase by 2026, accounting for 30% of total bit supply, with HBM4 and HBM4E combined expected to exceed 70% by 2027 [2]. Demand Drivers - HBM bit demand is forecasted to accelerate again in 2027, primarily driven by the Vera Rubin GPU and AMD MI400 [3]. - From 2024 to 2027, the CAGR for bit demand from ASICs, NVIDIA, and AMD is projected to exceed 50%, with NVIDIA expected to dominate demand growth [3]. - Sovereign AI demand is emerging as a key structural driver, with various countries investing heavily in national AI infrastructure to ensure data sovereignty and security [3]. Pricing and Cost Structure - Recent discussions around HBM pricing are influenced by Samsung's aggressive pricing strategy to capture market share in HBM3E and HBM4 [4]. - HBM4 is expected to have a price premium of 30-40% over HBM3E12Hi to compensate for higher costs, with logic chip costs being a significant factor [4]. Market Landscape - SK Hynix is expected to lead the HBM market, while Micron is likely to gain market share due to its capacity expansion efforts in Taiwan and Singapore [5]. - Micron's HBM revenue grew by 50% quarter-over-quarter, with a revenue run rate of $1.5 billion, indicating a stronger revenue-capacity conversion trend compared to Samsung [6]. Industry Impact - HBM is driving the DRAM industry into a five-year upcycle, with HBM expected to account for 19% of DRAM revenue in 2024 and 56% by 2030 [7]. - The average selling price (ASP) of DRAM is projected to grow at a CAGR of 3% from 2025 to 2030, primarily driven by the increasing sales proportion of HBM [7]. - Capital expenditures for HBM are expected to continue growing, as memory manufacturers focus on expanding capacity to meet rising HBM demand [7].

而Blackwell架构的AI芯片出货还在继续。今年下半年，采用该架构的GB300出货逐渐进入高峰期。当 地时间7月3日，云服务公司CoreWeave宣布，公司成为第一家为客户部署最新GB300 NVL72系统的 AI云服务提供商，并计划继续扩大部署。 在上周的英伟达股东大会上，英伟达CEO黄仁勋则谈到市场对算力的需求还在增长。他表示，推理需要 更多计算。仅在过去一年时间里，tokens（词元）生成就激增了50到100倍。为了满足这种指数级的 激增，大型人工智能工厂正在各地崛起。仅微软在上个季度就处理了超过100万亿的tokens，是一年前 的5倍。 黄仁勋还表示，公司在多个领域都有增长机会，人工智能和机器人技术是最大的两个机会，代表着数万 亿美元级别的增长机会。未来将有数十亿台机器人、数亿辆自动驾驶汽车和数十万个机器人工厂，它们 都将由英伟达的技术驱动。 有市场研究机构预计，英伟达将继续带动半导体IC产业规模增长。TrendForce集邦咨询梳理了去年全 球Fabless公司的营收，排除芯片以外的业务后发现，营收前10名的厂商中，英伟达营收增长率达 125%，其他公司的增长率最多只有21%，英伟达拉动了整个 ...

在上周的英伟达股东大会上，英伟达CEO黄仁勋则谈到市场对算力的需求还在增长。他表示，推理需要更多计算。仅在过去一年时间里，tokens（词元）生 成就激增了50到100倍。为了满足这种指数级的激增，大型人工智能工厂正在各地崛起。仅微软在上个季度就处理了超过100万亿的tokens，是一年前的5 倍。 黄仁勋还表示，公司在多个领域都有增长机会，人工智能和机器人技术是最大的两个机会，代表着数万亿美元级别的增长机会。未来将有数十亿台机器人、 数亿辆自动驾驶汽车和数十万个机器人工厂，它们都将由英伟达的技术驱动。 有市场研究机构预计，英伟达将继续带动半导体IC产业规模增长。TrendForce集邦咨询梳理了去年全球Fabless公司的营收，排除芯片以外的业务后发现，营 收前10名的厂商中，英伟达营收增长率达125%，其他公司的增长率最多只有21%，英伟达拉动了整个行业增长。该机构称，上半年半导体IC产业需求状况 很好，主要受惠于行业提前备货需求，再加上英伟达GB200、GB300下半年出货量增长，预计将推动2025年全球半导体IC产业规模增长19%左右。 截至周四收盘，英伟达股价上涨1.33%，股价159.34美元/股 ...

Market Performance - The US stock market experienced strong performance on July 3, with all three major indices rising. The S&P 500 index increased by 51.94 points, or 0.83%, closing at 6279.36 points. The Nasdaq Composite index climbed 207.97 points, or 1.02%, ending at 20601.10 points. The Dow Jones Industrial Average rose by 344.11 points, or 0.77%, closing at 44828.53 points. Both the S&P 500 and Nasdaq indices set new closing records, marking the S&P 500's seventh record close of the year and the Nasdaq's fourth record close of the year. Due to the public holiday, the US stock market closed early with relatively light trading volume [1]. Technology Sector - The technology sector was the main driver of the market's rise, with several leading tech stocks recording significant gains. Nvidia's stock price rose by 1.3%, reaching a historic high with a market capitalization of $3.89 trillion. The company briefly surpassed the $3.9 trillion market cap threshold, coming close to Apple's record for the highest global market capitalization. Amazon's stock increased by 1.59%, Microsoft by 1.58%, and Meta Platforms by 0.76%. Apple saw a 0.52% increase, while Alphabet rose by 0.5%. Tesla was an exception in the tech sector, closing down by 0.1% [3]. - Nvidia's strong performance was supported by multiple factors. OpenAI recently announced it would not adopt Google's TPU chips on a large scale and would continue to rely on Nvidia's GPUs and AMD's AI accelerators for its model training and inference work. OpenAI's reasoning was that these two chip manufacturers' products are "performance-validated" and have "existing supply agreements." This statement sent a positive signal to the market, indicating that Nvidia and AMD will remain core suppliers for OpenAI, potentially limiting Google's growth in the AI hardware market share [3]. Employment Data - The US Bureau of Labor Statistics reported that June non-farm payrolls exceeded market expectations, providing significant support for the stock market. In June, non-farm employment increased by 147,000 jobs, far surpassing analysts' expectations of 110,000, representing a 33% increase. The unemployment rate fell to 4.1%, better than the expected 4.3%. Average hourly earnings rose by 0.2% month-over-month and increased by 3.7% year-over-year, indicating a moderate wage growth trend that helps alleviate inflationary pressures [4]. - Employment growth showed structural characteristics, with government sector employment increasing by 73,000 jobs, primarily driven by state and local education positions. Healthcare added 39,000 jobs, and social assistance increased by 19,000 jobs. The federal government saw a reduction of 7,000 jobs due to layoffs. Additionally, employment data for the previous two months was revised upward, with April's figures adjusted from 147,000 to 158,000 and May's from 139,000 to 144,000, totaling an upward revision of 16,000 jobs [4]. - The strong employment data impacted expectations for the Federal Reserve's monetary policy. Before the data release, traders estimated a 25% probability of a rate cut in July. Following the report, market expectations for a short-term rate cut quickly diminished. The Chicago Mercantile Exchange's FedWatch tool indicated that the likelihood of a July rate cut fell to single digits, and the expectation for a 25 basis point cut in September decreased from 74% a week prior to 68% [4].

强劲就业报告提振市场 当地时间7月3日，美股三大股指全线收涨，道指涨0.77%报44828.53点，标普500指数涨0.83%报6279.35点，纳指涨1.02%报20601.1点。标普500指数、 纳指均创下收盘历史新高。 美国劳工统计局周四公布的数据显示，经季节性因素调整后，6月非农就业人数增加14.7万人，高于预期的11万人。失业率降至4.1%。就业人数稳步增长、 失业率下降，平均时薪环比增长0.2%，同比增长3.7%，表明通胀压力减弱。 政府部门就业人数大幅增加，增幅居所有类别之首，增加了7.3万人，原因是州政府和地方部门的就业大幅增加，尤其是与教育相关的工作岗位增加。联邦政 府仍受到马斯克政府效率部裁员的影响，损失了7000人。此外，医疗保健也很强劲，增加了3.9万人，而社会援助就业增加了1.9万人。 劳工统计局数据还显示，4月份非农新增就业人数从14.7万人修正至15.8万人；5月份非农新增就业人数从13.9万人修正至14.4万人。修正后，4月和5月新增 就业人数合计较修正前高1.6万人。 脑再生科技再度暴涨。 当地时间7月3日（周四），美股三大股指集体上涨，截至收盘，标普500指数涨0.83%， ...

Group 1 - Nvidia has regained its position as the world's most valuable company, surpassing Microsoft, but faces new challenges from OpenAI's shift towards Google's TPU chips for AI product support [1][3] - OpenAI's transition from Nvidia's GPUs to Google's TPUs indicates a strategic move to diversify its supply chain and reduce dependency on Nvidia, which has been the primary supplier for its large model training and inference [3][5] - The high cost of Nvidia's flagship B200 chip, priced at $500,000 for a server equipped with eight units, has prompted OpenAI to seek more affordable alternatives like Google's TPU, which is estimated to be in the thousands of dollars range [5][6] Group 2 - Google's TPU chips are designed specifically for AI tasks, offering a cost-effective solution compared to Nvidia's GPUs, which were originally developed for graphics rendering [8][10] - The TPU's architecture allows for efficient processing of matrix operations, making it particularly suitable for AI applications, while Nvidia's GPUs, despite their versatility, may not be as optimized for specific AI tasks [10][11] - The demand for inference power in the AI industry has surpassed that for training power, leading to a shift in focus among AI companies, including OpenAI, towards leveraging existing models for various applications [15]

在这个快速变化的时代，人工智能技术正以前所未有的速度发展，带来了广泛的机会 。 《AI日 报 》致力于挖掘和分析最新的AI概念股公司和市场趋势，为您提供深度的行 业 洞察和价 值 分 析。 A I 快 报 1. 据TrendForce集邦咨询最新研究，北美大型CSP目前仍是AI Server（服务器）市场需求扩 张主力，加上tier-2数据中心和中东、欧洲等主权云项目助力，需求稳健。 在北美CSP与OEM客 户需求驱动下，预估2025年AI Server出货量将维持双位数成长，然而因国际形势变化， TrendForce集邦咨询微幅下调今年全球AI Server出货量至年增24.3%。 2.知情人士称，"AI独角兽"Anthropic年化收入已达到每年40亿美元， 即每月3.33亿美元，较今 年年初增长了近四倍。 整理 | 美股研究社 3. OpenAI CEO Sam Altman就Meta近期大举挖角AI人才一事展开反击 。周一晚间，Altman 向OpenAI研究人员发出了一份措辞强硬的回应。回应中称："Meta的行为方式让人感觉有些厌 恶。" 周一，Meta CEO扎克伯格宣布了新成立的超级智能团队，其 ...

Core Insights - OpenAI currently has no plans to use Google's self-developed chips for its products despite initial testing of Google's Tensor Processing Units (TPUs) [1] - OpenAI has started renting Google's AI chips to meet its growing computational demands, marking its first significant use of non-NVIDIA chips [1] - OpenAI aims to reduce inference costs by leveraging Google's TPUs, which are expected to be a cheaper alternative to NVIDIA GPUs [1] - The company continues to actively use NVIDIA GPUs and AMD AI chips while also developing its own chips, with a key milestone of "tape-out" expected this year [1] Industry Dynamics - OpenAI has signed a contract to use Google Cloud services to address its increasing computational needs, indicating an unexpected collaboration between two major competitors in the AI field [2] - Despite the collaboration with Google, most of OpenAI's computational power still comes from GPU servers provided by emerging cloud service company CoreWeave [2]

Core Insights - The article emphasizes the ongoing evolution of AI terminals, highlighting the need for improvements in mobile AI functionalities while noting structural innovations in hardware such as optical, foldable screens, and fingerprint recognition. The recent surge in smart glasses sales is also mentioned, with a focus on the successful transition from AI glasses to AR glasses, as exemplified by Meta & Rayban AI glasses [3][4]. AI Terminal Development - AI glasses currently have limited interaction modes and functionalities, but the integration of AR features can significantly enhance user experience. The optical display module is expected to become a major component in AR glasses, with MicroLED and diffractive waveguides being the leading technologies [3]. Investment Opportunities - The long-term narrative for the AI industry remains strong, with companies like NVIDIA continuing to perform well. The rise of cloud vendors and breakthroughs in domestic computing power are expected to create diverse investment opportunities. Key sectors to focus on include servers, PCB, CPO, copper cables, power supplies, and liquid cooling, where domestic companies have established advantages [3][4]. Recommended Companies - Suggested companies for investment include: 1. Servers: Industrial Fulian, Huqin Technology 2. Computing Chips: Chipone, Cambricon, Haiguang Information 3. PCB: Huitian Technology, Shenghong Technology, Guanghe Technology, Shengyi Technology, Jingwang Electronics, Weier High 4. Copper/Optical Interconnect: Ruikeda, Bochuang Technology, Taicheng Light, Dongshan Precision 5. Power and Temperature Control: Hewei Electric, Zhongheng Electric, Magmi Tech, Shenxian Environment, Jianghai Co. 6. Brands and OEMs: Xiaomi Group, Yingshi Innovation, Goer Technology, Guoguang Electric 7. SOC: Lexin Technology, Hengxuan Technology, Xingchen Technology 8. Storage: Zhaoyi Innovation, Purang Co. 9. Distributors: Doctor Glasses, Kid King, Mingyue Lenses [4]. Market Trends - The article notes that the AI hardware and software sectors have seen significant stock price increases, with NVIDIA's stock rising by 45% and CoreWeave's by 195% since April 7. This reflects a broader trend of optimism in the AI market following NVIDIA's strong earnings report [17][18]. AI Chip Market Dynamics - The article discusses the increasing demand for ASICs as a key growth area in the AI chip market, with major cloud service providers like Google and Amazon ramping up their self-developed ASIC production. The global ASIC market is projected to grow from $6.5 billion in 2024 to $15.2 billion by 2033, with a compound annual growth rate of 12.8% [26][60]. Cloud Vendor Developments - Major cloud vendors are increasingly focusing on self-developed ASICs, with Google and Amazon leading the way. The article highlights that the market is shifting from NVIDIA's dominance to a more competitive landscape with multiple strong players emerging [60][61].

1. 6月30日(周一)美股盘前，美股三大股指期货齐涨。截至发稿，道指期货涨0.56%，标普500指数期货涨0.41%，纳指期货涨0.62%。 | ■ US 30 | 44,066.50 | 44,125.50 | 43,922.50 | +247.20 | +0.56% | | --- | --- | --- | --- | --- | --- | | ■ US 500 | 6,198.40 | 6,203.40 | 6,173.10 | +25.30 | +0.41% | | ■ US Tech 100 | 22,674.90 | 22,697.30 | 22,540.30 | +140.70 | +0.62% | 2. 截至发稿，德国DAX指数跌0.06%，英国富时100指数跌0.17%，法国CAC40指数涨0.03%，欧洲斯托克50指数跌0.13%。 | I | 德國DAX30 | 23,985.69 | 24,168.50 | 23,964.93 | -14.58 | -0.06% | | --- | --- | --- | --- | --- | --- | --- | | 淵 英國富時100 | ...