公众号记得加星标⭐️，第一时间看推送不会错过。 来源：内容 编译自 yahoo 。 Marvell Technology用 5nm 和 3nm 节点的先进 CMOS 技术，目前正转向 2nm 及以下节点，其中包括环绕栅极晶体管和 背面供电等创新。 这些因素吸引了超大规模企业积极利用这一技术。随着数据中心基础设施投资的快速增长，Marvell Technology 有望从中 获益。MRVL 预计，到 2028 年，其数据中心半导体的潜在市场规模将增至 940 亿美元。该公司预计，其加速定制计算产 品规模将达到 554 亿美元，2023 年至 2028 年的复合年增长率将达到 53%。 博通(Broadcom ) 的 AVGO 先进 3.5D XDSiP 封装平台专为提升用于 AI 加速器的定制 AI XPU 的性能和效率而设计。 博通的半导体部门（负责其定制硅片解决方案）在 2025 财年第一季度同比增长 11%。AMD 是定制硅片解决方案和 AI 加速器领域的另一家参与者，其半定制 SoC 产品和 Instinct 加速器为众多数据中心提供支持。 AMD 的可重构 Alveo 自适应加速器卡用于加速数据中心的 ...

Core Insights - Broadcom is experiencing significant growth in AI revenues, reporting a 46% year-over-year increase to $4.4 billion in Q2 of fiscal 2025, with AI networking revenues surging over 170% [1][10] - The company anticipates AI semiconductor revenues to reach $5.1 billion in Q3 of fiscal 2025, indicating a potential 60% year-over-year growth [2][10] - Broadcom's Ethernet-based networking portfolio, including Tomahawk switches and Jericho routers, is widely adopted by major hyperscalers like Google, Meta, and Microsoft, highlighting the importance of networking in AI workloads [3] AI Technology and Innovations - The launch of the next-generation Tomahawk 6 Ethernet switch, capable of 102.4 Tbps, aims to address networking bottlenecks in high-performance AI systems [4][10] - Innovations such as Cognitive Routing 2.0 and co-packaged optics are designed to enhance AI-scale fabrics, further solidifying Broadcom's position in the AI market [4] Competitive Landscape - NVIDIA remains a dominant player in the AI semiconductor market, offering high-performance GPUs and cloud solutions that drive scalable AI deployment [6] - Intel is advancing its AI strategy with new CPU and GPU architectures aimed at edge and data center workloads, indicating increased competition in the AI semiconductor space [7] Financial Performance and Valuation - Broadcom's stock has returned 18.8% year-to-date, outperforming the Zacks Electronics - Semiconductors industry, which grew by 15% [8] - The company is currently trading at a forward price-to-sales ratio of 18.11X, significantly higher than the industry average of 8.72X, suggesting potential overvaluation [11] - The Zacks Consensus Estimate for Q3 fiscal 2025 earnings is $1.66 per share, reflecting a 33.87% year-over-year growth [14]

AI算力需求持续增长，英伟达股价也不断创下新高。 美股7月9日开盘，英伟达股价一度上涨2.8%至164.42美元/股，市值首次突破4万亿美元，一举超过微软成为全球市值最大的科技公司。 英伟达也成为全球历史上首个实现这一市值水平的公司，超过了英、法、德等国家的股票总市值。此前，苹果创下过3.915万亿美元收盘市值纪录。 英伟达的崛起速度可谓惊人。特别是自2022年底ChatGPT推出以来，人工智能硬件和芯片的需求不断增长，英伟达从中获得了巨额利润。 2023年6月，英伟达市值首次突破1万亿美元，2024年2月突破2万亿美元，2024年6月突破3万亿美元，如今达到4万亿美元，增长速度远超苹果、微软等巨 头。 受DeepSeek模型及H20芯片对华出口限制，英伟达股价今年年初曾出现短暂大幅回调，股价一度跌破100美元。当时市场也不断有声音质疑英伟达的估值 合理性、未来可持续性等。 不过英伟达很快走出低谷，并强劲反弹。较今年4月的低点，英伟达股价已上涨89%。2023年以来，英伟达股价涨幅已超10倍。如今，英伟达的市值远超 其他芯片公司，约相当于39个英特尔、17.7个AMD。 得益于AI芯片市场需求的激增，以及英伟 ...

Welcome to our GTC Paris at VivaTech 2025 wrap-up. Hello everyone, I'm Dion Harris, Senior Director of HPC and AI Infrastructure Solutions at NVIDIA. Europe was the birthplace of the first two industrial revolutions and where modern manufacturing emerged.Now, the fourth industrial revolution has arrived: the age of AI. This year, we showcased something remarkable—the acceleration of AI across every corner of Europe. In his keynote, Jensen Huang shared how we're entering the era where agentic AI can understa ...

| Company | | Location | Number Of Chips | Type Of Chip | Status | Notes | | --- | --- | --- | --- | --- | --- | --- | | | | U.S. | 16,384 | H100 | Operating | Nvidia rents these servers for its DGX Cloud service | | awa | Amazon Web Services | Berwick, Pa. | Unknown | GPU | Planned | Data centers would be located | | | | | | | | next to a nuclear power plant | | | | Phoenix area | Unknown | GPU | Operating | | | | | U.S. | >200,000 | Trainium2 | Planned | AWS plans to build a Trainium | | | | | | | | clust ...

Core Viewpoint - MediaTek is entering a harvest period with its personal AI supercomputer strategy, collaborating with NVIDIA on the GB10 super chip, which is gaining traction among major PC brands and is set for mass shipment this month [1][2] Group 1: Product Development and Market Impact - The GB10 super chip is expected to significantly contribute to MediaTek's non-mobile business, enhancing its performance in the AI sector [1] - NVIDIA's CEO Jensen Huang announced the collaboration at CES, highlighting the GB10 chip as a key component of the world's smallest AI supercomputer, Project DIGITS [1] - The DGX Spark, developed in partnership with NVIDIA, offers unprecedented AI performance of 1,000 TOPS and supports AI models with up to 200 billion parameters [1][2] Group 2: Strategic Partnerships and Future Outlook - Major PC brands including Acer, ASUS, Dell, Gigabyte, HP, Lenovo, and MSI are set to supply DGX Spark products starting in July, which is seen as beneficial for both NVIDIA's AI market position and MediaTek's expansion into AI [2] - MediaTek acknowledges the uncertainties brought by tariff changes and is focused on strengthening collaborations with global supply chain partners to enhance adaptability [2] - The company maintains a positive long-term outlook on the pervasive trend of AI, supported by its solid financial position and ongoing investments in key technologies [2]

公众号记得加星标⭐️，第一时间看推送不会错过。 来源：内容 编译自 tomshardware 。 VideoCardz发现华硕的一份活动宣传邀请函显示，基于英伟达 GB10 Grace Blackwell 平台的实体系 统将于 7 月 22 日上市。华硕计划在当天重新推出其 Ascend GX10 迷你电脑。然而，该演示几乎没 有透露上市时间或实际价格，尽管英伟达此前曾提及过这两项信息。 VideoCardz发布的宣传单中写道："华硕 Ascent GX10 产品发布会：探索紧凑型华硕 Ascent GX10 的强大功能。了解其关键优势及其如何赋能人工智能开发。Nvidia 主题演讲嘉宾：聆听 Nvidia 专家 讲述 DGX Spark 软件堆栈的前沿功能，以及它们如何重新定义人工智能性能。" 描述清楚地表明华硕 Ascend GX10 系统超越了"迷你电脑"的范畴，并将该设备直接推向工作站和边 缘 AI 领域。 Nvidia 的GB10 Superchip系统级封装 (SiP) 集成了 Grace CPU，该 CPU 包含 10 个高性能 Arm Cortex-X925 核心，运行频率高达 3.90 GH ...

Group 1 - The core point of the article is the passage of the "Big and Beautiful Act," which significantly increases U.S. military spending to $1 trillion, raising questions about its effectiveness in revitalizing the military [1][3][5] - The act was passed with a narrow margin in both the Senate and House, reflecting significant controversy and debate surrounding its implications for military funding and strategy [3][5][6] - The act allocates $156.2 billion specifically for new military projects, with a total military budget approaching $960 billion for the upcoming fiscal year [5][6][7] Group 2 - The funding breakdown includes $70 billion for improving military personnel quality of life, $29.1 billion for shipbuilding, and $24 billion for missile defense systems, among other allocations [7][8] - The military's current procurement strategy is under scrutiny, as the number of weapons being purchased is decreasing despite rising budgets, indicating potential inefficiencies [6][22] - The article highlights a shift in military strategy under the Trump administration, moving away from previous policies and focusing on more pragmatic military projects [19][21][22] Group 3 - The article discusses China's naval advancements, particularly the recent dual aircraft carrier exercises, marking a significant development in naval capabilities and strategy [23][24][27] - The performance of China's aircraft carriers, particularly in terms of sortie rates, is compared favorably to U.S. naval operations, indicating a shift in naval power dynamics [29][30][34] - The article emphasizes the importance of these developments in the context of U.S.-China military competition, suggesting that the era of U.S. naval dominance is being challenged [41][42]

近日，Dylan Patel参加了Matthew Berman的播客访谈节目，对目前"硅谷AI大战"进行了深度而富有洞察的剖析。 Dylan Patel 是一位在半导体和人工智能领域有着深厚见解的专家。他创办了 SemiAnalysis，一家提供深入技术分析和市场洞察的机构。Patel 以其 对芯片产业、AI模型发展趋势、以及大型科技公司战略布局的独到见解而闻名。 Dylan Patel认为，近期Meta收购Scale AI并非看重其日渐"过气"的数据标注业务，而是为了Alexandr Wang及其核心团队。Wang的加入标志着扎克 伯格在AI战略上的重大转变，从之前的"AI很好，但AGI不会很快实现"转变为全力追逐"超级智能"，因为他意识到Meta在该领域已经落后，需 要"迎头赶上"。 在设备端AI方面，Dylan Patel表达了悲观的看法。他认为，尽管设备端AI在安全和低延迟方面有优势，但消费者对价格的敏感度更高，更倾向于 使用免费的云端AI服务。他预测，设备端AI的应用将主要限于低价值、轻量级的任务，例如在可穿戴设备上进行图像识别或手部追踪，而复杂且 有价值的AI功能仍将依赖于云端。苹果自身也在建设大 ...

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]