如果您希望可以时常见面，欢迎标星收藏哦~ 来源：本文综合自网络 ，谢谢。 欧洲面临着艰难的选择——为"主权"而战、与中国亲近，还是与华盛顿合作。 据报道，特朗普政府准备对半导体征收大范围关税。目标不仅限于中国或外国成品芯片。潜在的 所谓"零部件关税"将把设备内部外国制造芯片的价值加总，并对成品设备征收关税。 且不论如此复杂的方案是否切实可行（严肃的分析师对此表示怀疑），关税可能会损害欧洲的利 益，减少其自身尖端芯片的销售，并减缓欧洲大陆的芯片投资。欧洲最好的应对措施并非通过芯 片补贴进行反击，而是与中国合作，并在必要时对一系列美国数字产品征税进行反击。 我们先来看看潜在的直接影响。欧洲对美国的芯片出口几乎不会受到影响。欧洲仅占全球芯片产 量的8%，而且大部分都销往欧洲汽车行业，而不是在亚洲组装的消费电子产品。 欧洲真正的优势——以及其脆弱性——在于生产制造尖端芯片所必需的机器。领先的荷兰光刻设 备制造商ASML将受到影响。尽管该公司计划将"大部分"关税成本转嫁给美国客户，但由于贸易 战影响的不确定性，该公司已经预测订单将放缓。由于非人工智能芯片产能过剩，包括英特尔、 台积电和三星在内的ASML客户正在推迟设备 ...

Core Viewpoint - The AI chip market is shifting focus from training to inference, with companies like Graphcore, Intel, and Groq adapting their strategies to capitalize on this trend as the training market becomes increasingly dominated by Nvidia [1][6][12]. Group 1: Market Dynamics - Nvidia remains the leader in the training chip market, with its CUDA toolchain and GPU ecosystem providing a significant competitive advantage [1][4]. - Companies that previously competed in the training chip space are now pivoting towards the more accessible inference market due to high entry costs and limited survival space in training [1][6]. - The demand for AI chips is surging globally, prompting companies to seek opportunities in inference rather than direct competition with Nvidia [4][12]. Group 2: Company Strategies - Graphcore, once a strong competitor to Nvidia, is now focusing on inference, having faced challenges in the training market and experiencing significant layoffs and business restructuring [4][5][6]. - Intel's Gaudi series, initially aimed at training, is being repositioned to emphasize both training and inference, with a focus on cost-effectiveness and performance in inference tasks [9][10][12]. - Groq has shifted its strategy to provide inference-as-a-service, emphasizing low latency and high throughput for large-scale inference tasks, moving away from the training market where it faced significant barriers [13][15][16]. Group 3: Technological Adaptations - Graphcore's IPU architecture is designed for high-performance computing tasks, particularly in fields like chemistry and healthcare, showcasing its capabilities in inference applications [4][5]. - Intel's Gaudi 3 is marketed for its performance in inference scenarios, claiming a 30% higher inference throughput per dollar compared to similar GPU chips [10][12]. - Groq's LPU architecture focuses on deterministic design for low latency and high throughput, making it suitable for inference tasks, particularly in sensitive industries [13][15][16]. Group 4: Market Trends - The shift towards inference is driven by the lower complexity and resource requirements compared to training, making it more accessible for startups and smaller companies [22][23]. - The competitive landscape is evolving, with a focus on cost, deployment, and maintainability rather than just computational power, indicating a maturation of the AI chip market [23].

Core Insights - Ampere Computing has introduced the AmpereOne M Arm server CPU, featuring a 12-channel memory configuration, enhancing its capabilities to compete with AMD EPYC 9005 Turin and Intel Xeon 6900P [1][3][6] Product Overview - The AmpereOne M is set to be widely available by Q2 2024, with shipments expected in Q4 2024 [3][6] - The new CPU architecture retains a relatively small SKU stack, with naming conventions indicating core count and clock speed [10] - Key specifications include support for 96 to 192 custom Arm cores, 12-channel DDR5-5600 memory, and a maximum addressable capacity of 3TB [18][20] Technical Specifications - The AmpereOne M supports 12-channel memory with a single DIMM per channel, and features a system-level cache of 64MB [18] - It includes 96 PCIe Gen5 lanes, allowing for bifurcation down to x4, and supports up to 24 discrete attached devices [18][20] - The memory configuration is limited to 1 DPC (DIMM per channel), which may impact performance compared to competitors [17] Market Context - The acquisition of Ampere by SoftBank suggests a potential shift in focus for the company, raising questions about its customer and partner visibility [21] - The current landscape indicates that Ampere may dominate the enterprise server Arm market, especially as competitors like NVIDIA Grace are not ideal for enterprise applications [21]

如果您希望可以时常见面，欢迎标星收藏哦~ 来源：本文编译自路透社 ，谢谢。 芯片制造商 Wolfspeed 对 2026 年营收的预测低于华尔街预期，并于周四公布第三季度营收下降 7%，导致其股价在盘后交易中下跌 11%。 该公司使用碳化硅（一种比传统硅更节能的材料）制造芯片，但受到电动汽车普及速度低于预期以 及新关税导致汽车零部件价格上涨的影响。 这些因素导致客户推迟产品发布，从而削弱了对 Wolfspeed 产品的需求。 更广泛的宏观经济挑战，包括高利率和资本成本上升，也推迟了工业和能源领域的投资周期，这可 能会进一步给订单活动带来压力。 该公司预计 2026 年营收为 8.5 亿美元，低于分析师预期的 9.587 亿美元。 然而，在特朗普政府呼吁立法者废除《芯片法案》下的联邦资金后，拜登时代承诺为国内芯片制造 业提供补贴的法案的未来仍不确定。这导致Wolfspeed的股价在3月份下跌了一半，跌至27年来的 最低点。 Wolfspeed 报告第三季度每股净亏损为 72 美分，而预计每股亏损为 82 美分。 Wolfspeed，还好吗？ 如果您不熟悉，Wolfspeed 生产的是碳化硅。这种对普通硅的改进使 ...

Core Viewpoint - The article discusses the evolution and significance of Moore's Law in the semiconductor industry, emphasizing its historical context, current challenges, and future prospects for technology advancements in integrated circuits [3][5][6]. Group 1: Historical Context of Moore's Law - Moore's Law originated from Gordon Moore's observation in 1965 that the number of components on integrated circuits would double approximately every two years, leading to exponential growth in circuit complexity [5][15]. - The early development of the integrated circuit industry was driven by military defense needs, with significant investments from the U.S. Air Force in the 1960s [6][9]. - Predictions made by industry leaders in the 1960s indicated that integrated circuits would become cost-competitive with traditional circuits, with estimates suggesting that by 1973, integrated circuits could be priced at one-third to two-thirds of traditional circuit costs [9][10]. Group 2: Current Challenges and Future of Moore's Law - There is ongoing debate about the "death" of Moore's Law, with concerns about potential technological barriers. However, the semiconductor industry has consistently overcome such challenges through innovation [6][22]. - The article highlights the transition from traditional planar transistors to advanced architectures like FinFET and GAA (Gate-All-Around) transistors, which enhance performance and efficiency [35][37]. - The semiconductor industry aims to continue reducing power consumption and costs while increasing performance and integration density, with a focus on maintaining the trajectory of Moore's Law [27][41]. Group 3: Technological Innovations - Innovations such as strained silicon and high-k materials have been pivotal in enhancing transistor performance and enabling further miniaturization [31][32]. - The introduction of 3D transistor architectures, such as FinFET and GAA, represents a significant advancement in transistor design, allowing for better control and efficiency [35][37]. - Future technologies, including CFET and FFET, are anticipated to further push the boundaries of transistor scaling and performance, continuing the legacy of Moore's Law into the next decade [39][41].

如果您希望可以时常见面，欢迎标星收藏哦~ 来源：本文 综合自中芯国际 ，谢谢。 昨夜晚间，中芯国际发布了2025年第一季的财务数据，数据显示，在一季度，公司整体实现销售 收入22.47亿美元，环比增长1.8%；毛利率为22.5%，环比大致持平；产能利用率上升至89.6%， 环比增长4.1个百分点。 在毛利方面，公司2025年第一季录得毛利为505.9百万美元，2024年第四季毛利为499.0百万美 元 ， 同 比 增 长 111.0% 。 如 下 图 所 示 ， 在 经 营 利 润 方 面 ， 公 司 在 2025 年 一 季 度 获 得 了 同 比 大 增 12766.6%，环比增加也达到44.3%。 从图表中可以看到，归属到公司拥有人的利润，也环比大增74.8%，同比增加幅度更合适高达 161.9%。 | | 2025 年 | 2024 年 | 季度比較 | 2024 年 | 年度比較 | | --- | --- | --- | --- | --- | --- | | | 第一季度 | 第四季度 | | 第一季度 | | | 收入 | 2.247.201 | 2.207.281 | 1.8% | 1.75 ...

Core Insights - Apple is making significant progress in developing a custom chip for its upcoming smart glasses, indicating an acceleration in its product development efforts to compete with Meta's offerings [1][2] - The chip is designed for low power consumption and will control multiple cameras in the glasses, with production expected to start by the end of next year or in 2027 [1][2] - Apple is also exploring non-AR glasses that utilize AI to assist users, aiming to enter a market where Meta has already established a presence [2] Group 1: Smart Glasses Development - Apple is developing a smart glasses chip based on technology used in the Apple Watch, which has lower power consumption compared to components in iPhone, iPad, and Mac [1] - The company plans to start mass production of the smart glasses processor by the end of next year or in 2027, suggesting a potential market launch within the next two years [1] - Apple is conducting user research on the smart glasses concept, indicating a commitment to understanding consumer needs before launch [2] Group 2: AI and Semiconductor Innovations - Apple is expanding its semiconductor capabilities, developing chips for AirPods and Apple Watch that include camera functionalities, with a target completion date around 2027 [3] - The company is also working on AI server chips, named "Baltra," which will enhance the processing of Apple Intelligence requests, with a completion goal set for 2027 [4][5] - New Mac chips, including potential M6 and M7 processors, are in development, with plans to introduce the M5 processor for iPad Pro and MacBook Pro by the end of this year [3][5]

Core Insights - The xAI Memphis Supercluster has reached full operational capacity, utilizing 150 MW from the Tennessee Valley Authority (TVA) and an additional 150 MW from Megapack batteries for backup power [1][2] - The Colossus supercomputer, equipped with 100,000 Nvidia H100 GPUs, was deployed in just 19 days, a process that typically takes four years [1][11] - Future expansions aim to double the GPU count to 200,000, with plans to eventually reach 1 million GPUs, significantly increasing the power and capabilities of the supercomputer [3][7] Power Supply and Infrastructure - The first phase of the project can now operate entirely on TVA power, which sources about 60% of its energy from renewable resources [2] - A second substation is expected to be operational by fall 2023, increasing total power capacity to 300 MW, sufficient to power 300,000 homes [2] - Initial reports indicated the presence of 14 gas turbines on-site, with some residents noting over 35 turbines, raising concerns about local energy supply [1] Technological Advancements - Colossus is designed to push the boundaries of AI research, focusing on training large language models and exploring applications in autonomous vehicles, robotics, and scientific simulations [6][13] - The upcoming Nvidia Blackwell H200 GPUs promise significant performance improvements, potentially up to 20 times faster than the H100 GPUs, although delivery has faced delays due to design issues [7][8] - The infrastructure includes advanced cooling systems to manage the heat generated by the high-density GPU setup, which is critical for maintaining performance [14][15] Competitive Landscape - The investment in Colossus positions xAI to compete effectively against major players like Google, Microsoft, and OpenAI in the AI research space [15] - The ability to rapidly train AI models could lead to breakthroughs that were previously limited by computational constraints, enhancing xAI's research capabilities [15] - Concerns have been raised regarding the geopolitical implications of foreign ownership of advanced AI technologies, particularly in non-research applications [16]

如果您希望可以时常见面，欢迎标星收藏哦~ 来源：本文 编译自 chousun ，谢谢。 随着越来越多的顶尖理工科学生放弃芯片相关专业，转而投身医学领域，韩国正面临半导体行业年 轻人才的流失。在全球对半导体创新需求不断增长的背景下，这种转变正在重塑韩国的高科技劳动 力队伍。 全国各地的大学都报告称，越来越多的学生退出了半导体项目，其中包括曾经将人才直接输送到三 星电子和 SK 海力士等大公司的精英研究生课程。 教育界人士警告称，政府扩大医学院招生的计划正在加速这一趋势，吸走韩国最关键行业之一的人 才。 其中一个案例涉及一名研究生，仅被确认为 A 先生，他在首尔一所私立大学攻读半导体工程综合 硕士和博士课程仅一年后就离开了。 他从同一所学校的电子工程系毕业后，就读研究生，目标是加入一家大型半导体公司的研发团队。 但去年2月，他提交了退学申请，改变了学习方向。 "我听其他学生说，他现在正在准备读医学院，"他的学术导师说。"即使在半导体这种就业几乎有 保障的领域，年轻人也选择离开，去寻找更有保障的未来。" 据政府运营的高等教育门户网站"Allimi"称，离开半导体工程研究生课程的学生人数从2023年的 17人增加到去 ...

以下文章来源于东芝半导体芯资讯 ，作者芝子 东芝半导体芯资讯 . 东芝半导体资讯速递平台，为您呈现东芝半导体最新企业要闻、重磅新品动态、前沿技术解决方案以及 各类线上线下活动资讯，助您第一时间掌握企业动态及行业脉搏，让我们携手共同开启智能生活应用的 无限可能，探索科技赋能下的美好未来。 "东芝半导体芯资讯"公众号官宣 您的"芯"伙伴已上线！ 无论是东芝在存储芯片领域的创新突破， 还是在汽车、工业、消费电子等 多领域 的产品 应用新成果， 它将化身您的专属信息中枢， 成为您在工作生活中的知"芯"伙伴。 在这里， 将及时为您推送东芝半导体的 最新 企业 要闻、 重磅新品动态以及前沿技术解决方案， 更有行业展会、线上研讨会等精彩活动， 亲爱的伙伴们， 今天，我们带着一份特别的"芯"意而来， "东芝半导体芯资讯"公众号正式官宣上线！ 都能为您第一时间解锁， 助您精准捕捉每一次"芯"变化。 初次见面，好礼奉上 点击下方按钮，跳转关注公众号，开启你我信息传递的"芯"桥梁。 参与条件 点击下方卡片，即刻关注！ 丰富的内容类型 企业新闻、新品推荐、前沿技术、案例分享、技术讲堂、精彩视频、展会资讯、线上研讨 会、互动活动等。 ...