在过去的美好时光里，网络只是连接少量本地计算机。但时代变了。在人工智能主导的世界里，关键 在于协调数万台服务器的活动，以训练大型语言模型，并且不造成任何通信延迟。现在，有一个经过 优化的架构可以实现这一点。Cornelis Networks表示，其 CN500 网络结构可最大限度地提高人工智 能性能，支持部署多达 50 万台计算机或处理器（比现在高出一个数量级），并且不会增加延迟。 继以太网和InfiniBand之后，这项新技术为网络领域带来了第三款重要产品。它旨在使人工智能和高 性能计算机（HPC，即超级计算机）能够以更高的效率实现更快、更可预测的完成时间。Cornelis 声 称，在 HPC 方面，其技术优于2022 年推出的InfiniBand NDR版本，每秒传递的消息数量是后者的 两倍，延迟降低了 35%。在人工智能应用方面，与基于以太网的协议相比，其通信速度提高了六 倍。 以太网长期以来一直是局域网（LAN）的代名词。软件补丁使其通信协议经受住了时间的考验。 InfiniBand 的发明是一项改进，但它的设计目标仍然是相同的：连接少量本地设备。"这些技术发明 之初，与并行计算毫无关联，"宾夕法尼亚 ...

Group 1 - The core viewpoint of the articles highlights the strengthening of the military industry amid geopolitical tensions and technological advancements, with a positive outlook for the sector driven by both external pressures and internal growth dynamics [2] - The CN5082.SZ aerospace and defense index saw a 0.25% increase, with notable gains from companies such as Guorui Technology (up nearly 6%) and Yaguang Technology (up over 3%) [1] - The recent adjustment of the CN5082.SZ index included the addition of five key stocks, such as AVIC Chengfei and China Shipbuilding Emergency Rescue, while five others were removed [1] Group 2 - The military industry is expected to benefit from the changing global military development environment and strong internal growth, with increasing military demand enhancing the strategic position of the sector [2] - The military industry demonstrates stronger anti-cyclicality compared to traditional sectors, making it more resilient amid global economic fluctuations [2] - Short-term drivers for the military industry include accelerated technological iterations and approaching key strategic milestones, which are likely to maintain an upward trend in industry prosperity [2]

如果您希望可以时常见面，欢迎标星收藏哦~ 来源：内容 编译自 theregister 。 在英特尔将其Omni-Path互连技术剥离给Cornelis Networks五年后，这家公司终于准备好凭借其 400Gbps CN5000系列交换机和网卡（NIC）正面对抗老对手Nvidia的InfiniBand技术。 这一次，Cornelis的目标不仅限于超级计算机和高性能计算（HPC）集群，它还希望借助更高的性 价比切入AI热潮，与Nvidia展开竞争。 对于那些早已将Omni-Path抛诸脑后的读者，以下是快速回顾：Omni-Path最初由英特尔于2015年 开发，是一种无损互连技术，在许多方面与Nvidia的InfiniBand网络相似，主要面向高性能计算应 用。 第一代Omni-Path交换机提供4.8Tbps带宽，共有48个100Gbps端口，曾部署于多个超级计算平 台，如洛斯阿拉莫斯国家实验室的Trinity系统和美国能源部的Cori系统。 但 到 了 2019 年 ， 英 特 尔 放 弃 了 该 项 目 ， 并 于 2020 年 9 月 将 该 业 务 部 门 剥 离 成 立 Cornelis Netw ...

Core Insights - AbbVie has received FDA approval for Emrelis, marking a significant advancement in its oncology franchise, specifically for treating non-squamous non-small cell lung cancer (NSCLC) with high c-Met overexpression [1][2] Group 1: Product Development and Market Position - Emrelis is AbbVie's first internally developed solid tumor drug and its first lung cancer therapy, providing a unique advantage as the only approved treatment for this patient population [2][8] - c-Met overexpression is present in 25% of advanced EGFR wild-type NSCLC patients, with approximately half of these patients exhibiting high c-Met overexpression, which is associated with poor prognosis [2] - AbbVie anticipates starting to record sales for Emrelis in Q3 2025, with estimated sales of $28 million for the current year [3] Group 2: Competitive Landscape - While AbbVie has a first-mover advantage with Emrelis, competitors like AstraZeneca and Regeneron Pharmaceuticals are also developing drugs targeting similar indications [4][8] - Regeneron is conducting mid-stage studies for its investigational ADC, davutamig, which targets two distinct MET epitopes [5] - AstraZeneca is in phase I studies for its investigational ADC, AZD9592, aimed at c-MET-driven solid tumors, although it currently lags behind AbbVie in this specific indication [6] Group 3: Financial Performance and Valuation - AbbVie shares have outperformed the industry year to date, reflecting positive market sentiment [7] - The company's shares trade at a price/earnings (P/E) ratio of 14.37, slightly below the industry average of 14.95, but higher than many other large drugmakers [10] - The Zacks Consensus Estimate for AbbVie's 2025 earnings has slightly decreased from $12.30 to $12.28 per share, while the estimate for 2026 has increased from $13.97 to $14.05 [11]

| 航空航大 | | --- | | 马斯克称 SpaceX 今年收入将超 155 亿美元，明年或超 NASA 整年预算 | | 日本 ispace 月球着陆器"韧性"号 6 月 6 日尝试着陆 . | | 俄航天集团: 俄储蓄银行的 GigaChat 神经网络助手或将于今秋引入国际空间站 14 | | 能源矿产 . | | 特朗普签署行政令 将进口钢铝关税提高至 50% | | 德国汽车业界警告中国稀土出口限制或致停产 … | | 主要产油国继续增产 油价下跌概率加大 | | 国际社会热点 | | --- | | 特朗普关税冲击美国与全球 OECD 再度下调全球经济预测 3 | | 白宫称中美领导人本周或进行通话,中国外交部回应 | | 中国商务部部长王文涛部长会见世贸组织总干事伊维拉 . | | 人工智能 . | | 日本计划借助 AI 加强知识产权竞争力 | | 韩国总统李在明:将大规模投资人工智能和芯片 | | 中国深圳: 人工智能终端技术攻关最高可资助 2000 万 . | | 微软将在瑞士投资 4 亿美元建设 AI 基础设施和云计算 | | Cornelis Networks 推出新技术，旨在加速 ...

Core Viewpoint - Cornelis Networks has launched the CN5000 series, a 400Gbps Omni-Path technology aimed at enhancing AI and HPC performance, with plans for future UltraEthernet integration [1][19]. Group 1: Product Overview - The CN5000 series is designed for AI and HPC applications, supporting deployments of up to 500,000 endpoints and providing advanced lossless data transmission and congestion avoidance features [19][20]. - The CN5000 utilizes Omni-Path architecture, which includes credit-based flow control and dynamic fine-grained adaptive routing to ensure high performance and reliability [20][22]. - The series includes both air-cooled and liquid-cooled switch options, addressing the cooling needs of large AI companies [5][22]. Group 2: Performance Metrics - The CN5000 demonstrates superior performance in HPC workloads, achieving up to 30% faster execution in computational fluid dynamics and climate modeling compared to InfiniBand NDR [22]. - For AI applications, the CN5000's collective communication speed is six times faster than RoCE, enhancing the efficiency of large language model training [19][22]. Group 3: Future Developments - The upcoming CN6000 series, set for release in 2026, will integrate Omni-Path with Ethernet support, doubling bandwidth capabilities [16][21]. - The CN7000 series, expected in 2027, will combine UltraEthernet standards with Omni-Path architecture, targeting the most demanding AI and HPC environments [17][21]. Group 4: Market Positioning - Cornelis Networks aims to provide a cost-effective alternative to NVIDIA's InfiniBand, capitalizing on the growing demand for high-performance networking solutions in AI and HPC sectors [18][19].

中信建投指出，军工行业配置方面建议围绕三条投资主线，一是传统军工方向，建议重点关注有订单修 复预期、有业绩支撑的航发产业、船舶产业以及航空航天产业；二是新域新质方向，新式战争方面我们 建议关注围绕低成本、智能化、体系化三大特点的行业，包括低成本精确制导弹药、无人系统以及新一 代智能化作战底座；新质生产力方面，建议关注应用市场空间广阔、行业处于快速成长期、国产化率较 低的行业，主要为商业航天、低空经济、大飞机；新技术方面，建议关注MEMS器件、增材制造以及陶 瓷基复合材料。三是改革出海方向，我们建议关注有资产整合预期、军贸市场具有竞争力的相关公司。 5月最后一个交易日，A股早盘走低，军工板块局部拉升。相关ETF中，截至发稿，航空航天ETF天弘 （159241）微跌0.2%，换手率超8%，成交额突破1500万元，盘中交投活跃。成分股中，航发科技涨超 8%冲击二连板，航发控制、航发动力、国科军工等多股跟涨。 值得一提的是，航空航天ETF天弘（159241）昨日上市首日收涨2.21%，当日成交额达1.82亿元，在同 标的产品中排名第一。 航空航天ETF天弘（159241）跟踪国证航天航空行业指数（代码CN5082，简 ...

Sameer Desai Vice President, Corporate Development & Investor Relations TTM Technologies, Inc. +1 714 327 3050 sameer.desai@ttmtech.com Technical Inquiries Mark Bowyer Director, Business Development, RF&S BU TTM Technologies, Inc. +1 315 278 5420 mark.bowyer@ttm.com 50Ω to 100Ω broadband balanced to unbalanced transformers: Our new ultra-small broadband transformers XMB0465Z1-50100G and XMB0220K1-50100G are designed to convert single-ended signals to differential ports with an exceptional common mode reject ...

来源 | 新智元 曾被寄予厚望的苹果AI，似乎已被这个飞速迭代的时代淘汰了。 Siri，作为曾经的语音助手先驱，光环逐渐暗淡。苹果AI的努力在谷歌OpenAI等巨头夹击之下，又显得力不从心。 如今，苹果正通过拆分AI团队，试图重塑组织架构，在这场鏖战中扳回一局。 苹 果 A I 六 年 沉 浮 2018年之前，苹果AI工作分散在公司各个部门。 Siri团队，以及其他与苹果iOS操作系统相关的AI小组，隶属于「软件工程部门」。 硬件团队有自己的AI成员，专注于研发苹果处理器神经引擎等组件。自动驾驶汽车项目，则由一些硬件高管负责。 此外，服务和运营部门也有一些AI相关的团队。 这里，没有专门的iPhone、iPad或Mac团队，这些产品均由产品硬件、软件、服务部门共同开发。 虽看似不合理，但这种按职能，而非业务单元组织的模式，是苹果一直以来的传统。 随着AI在消费设备中的重要性日渐凸显，分散的团队结构逐渐开始拖后腿。Siri从开创性产品逐渐被淘汰，远远落后于亚马逊的Alexa和谷歌助 手。 当时，苹果高管认为，「AI团队组织分散，对公司来说弊大于利」。 为了扭转局面，苹果于2018年从谷歌挖来了AI大咖John ...

Core Viewpoint - Lenovo is positioning itself in the AI era by launching the "Tianxi Personal Super Intelligent Agent" integrated into various AI terminals, aiming to enhance user experiences across work, study, and daily life scenarios [1][3]. Group 1: Product Launches - Lenovo introduced multiple AI terminal products featuring the Tianxi Personal Super Intelligent Agent, including the ThinkPad X1 Carbon 2025 Aura AI version, which boasts a thickness of 14.37mm and a battery life of up to 18 hours [4]. - The ThinkPad T14p 2025 AI version is equipped with Intel's new Ultra 9 processor, achieving 99 TOPS of local computing power, enhancing coding and data analysis efficiency [5]. - The ThinkCentre neo Ultra 2025 AI version is designed for high-load tasks, capable of handling complex data computations and model training [5]. Group 2: AI Applications and Features - The Think AI terminals incorporate five key applications: AI search for direct results, multimodal AI translation, a knowledge base for managing materials, AI tuning for device control, and AI services for image diagnosis and recommendations [3]. - The thinkplus meeting solutions include the T86 Pro meeting tablet with 4K camera and voice localization technology, enhancing collaboration and meeting documentation [6]. Group 3: Legal AI Strategy - Lenovo launched the "Lenovo Legal Intelligent Agent Strategy," which includes eight functional modules aimed at improving efficiency in legal work, such as dispute resolution and compliance management [7]. - The first product under this strategy is the "Lenovo Fasi AI Legal Assistant," developed in collaboration with Turing Law, designed to support lawyers throughout the litigation process [7][8].