Core Viewpoint - AMD's CDNA 4 architecture represents a moderate update over CDNA 3, focusing on enhancing matrix multiplication performance for low-precision data types, which are crucial for machine learning workloads [2][26]. Architecture Overview - CDNA 4 maintains a similar system-level architecture to CDNA 3, utilizing a large chiplet setup with eight compute dies (XCD) and a memory-side cache of 256 MB [4][20]. - The architecture employs AMD's Infinity Fabric technology for consistent memory access across multiple chips [4]. Performance Comparison - The MI355X GPU, based on CDNA 4, features a clock speed of 2.4 GHz and 256 cores, compared to MI300X's 304 cores at 2.1 GHz, indicating a slight reduction in core count but improved clock speed [5]. - MI355X offers 288 GB of HBM3E memory with a bandwidth of 8 TB/s, surpassing Nvidia's B200, which has a maximum capacity of 180 GB and bandwidth of 7.7 TB/s [25]. Matrix and Vector Throughput - CDNA 4 has rebalanced execution units to focus on low-precision matrix multiplication, doubling matrix throughput per compute unit (CU) in many cases [6][39]. - The architecture supports new low-precision data formats, significantly enhancing AI performance, with matrix core improvements leading to nearly four times the computational throughput for low-precision formats [46][47]. Local Data Sharing (LDS) Enhancements - CDNA 4 increases the Local Data Share (LDS) capacity to 160 KB and doubles the read bandwidth to 256 bytes per clock, improving data locality for matrix multiplication routines [14][48]. - The architecture introduces new instructions for reading transposed LDS, optimizing memory access patterns for matrix operations [18]. Memory Hierarchy and Cache - The memory hierarchy includes a shared 4 MB L2 cache and a 32 KB L1 vector cache per CU, with enhancements for caching non-coherent data from DRAM [49][50]. - The Infinity Cache remains at 256 MB, providing high bandwidth and supporting the increased memory demands of modern AI workloads [53]. Chiplet Architecture - The CDNA 4 architecture continues to leverage a chiplet-based design, allowing for independent evolution of each chiplet for improved performance and manufacturability [35][36]. - Each XCD contains 36 compute units, organized into arrays, with a focus on maximizing yield and operational frequency [39]. System Communication and Expansion - The architecture includes eight AMD Infinity Fabric links, with improved speeds of up to 38.4 Gbps, enhancing communication bandwidth within server nodes [63]. - The design supports both direct compatibility with previous generations and progressive improvements for high-performance systems [62]. Conclusion - AMD's CDNA 4 architecture builds on the success of CDNA 3, focusing on optimizing performance for machine learning workloads while maintaining a competitive edge against Nvidia [26][27].

Core Viewpoint - The article discusses the current state and future directions of photomask manufacturing, emphasizing the importance of curved masks and advanced computational tools in extending the viability of non-EUV lithography technologies [1][3][4]. Group 1: Innovations in Photomask Technology - The use of curved photomasks is a significant innovation that leverages current writing technologies to create complex shapes previously unattainable [3]. - Advanced computational tools, such as Mask Process Correction (MPC) and high-level simulations, are increasingly used in the mask design flow, reducing the need for expensive experiments and pushing technological limits [3][6]. - The evolution of variable shape beam (VSB) writing technology to multi-beam writing technology has made curved mask shapes feasible without increasing writing time or costs [5]. Group 2: Challenges and Infrastructure Needs - There is a substantial need for infrastructure development to support the complexity of curved shapes, as traditional rectangular descriptions are simpler to manage [8]. - The transition to curved processes is seen as an exception rather than the norm, impacting economics and infrastructure, particularly in the reliance on GPU-based computing [9]. - Measurement technologies must evolve to handle the complexities of curved shapes, requiring higher resolution and faster measurement tools [11]. Group 3: EUV Masking Issues - EUV masks face challenges such as lower durability compared to 193i masks, necessitating frequent replacements that increase costs and complexity [13]. - The performance of EUV pellicles is currently suboptimal, leading to significant wafer throughput losses due to energy loss during transmission [13][15]. - The balance between using pellicles and the associated costs is contingent on the specific use case, with larger, high-value chips benefiting more from pellicles than smaller, redundant designs [16]. Group 4: Future Directions and Research - Research is ongoing into alternative materials for pellicles, such as carbon nanotube films, which could address current limitations but are not yet in mass production [17]. - The industry is exploring ways to improve the durability and transmission rates of EUV pellicles, which could lead to broader applications if successful [15][16].

弘则科技- AI 应用调研 20250616 摘要 公司于 2021 年获得项目自主审批权，加速了人工智能项目落地，此前 受制于集团公司对百万以上项目的严格审批。这一政策转变源于集团层 面统一规划的推进困难，鼓励有实力的分子公司先行探索。 传统 ERP 系统已无法满足人工智能时代的需求，需要在边缘侧增加自动 感知和数据感知能力，以实现中心侧的智能化管理。公司并行运行机器 学习和人工建模，专家经验在短周期设备分析中仍具优势。 传统地面监控系统误报率高，通过逻辑建模筛选误报，提高精确报警率， 减少现场检查工作量。机器学习通过专家知识自动学习，处理大量样本， 实现设备状态的提前预测和预防。 公司通过综合传感器数据预测设备状态，并根据不同类型的场景预设构 建相关模型。智慧运行系统告警后联动生产管控系统，由现场维护人员 或远程操作进行响应，确保及时响应。 经过对多家供应商的调研，最终选择了新环科技和第四范式，主要考虑 成本因素。第四范式因其子公司在水电领域有丰富经验，能结合人工智 能技术提供清晰解释而被选定。 Q&A 贵公司最早是如何决定使用主动学习技术的？这个决策过程是怎样的？ 我们从 2022 年开始考虑智慧企业建 ...

以下文章来源于硬氪 ，作者黄楠 硬氪 . 专注全球化、硬科技报道。36kr旗下官方账号。 已同多家头部机器人厂商展开深入合作。 文 ｜ 黄楠 编辑 ｜ 袁斯来 来源｜ 硬氪（ID：south_36kr） 封面来源 ｜企业供图 硬氪获悉，深圳留形科技有限公司（以下简称「留形科技」）近日完成数千万元Pre-A轮融资，本轮投资方包括弘毅投资等，资金 将主要用于核心零部件定制生产、产品规模化交付及市场拓展。此前，公司曾获得真格基金种子轮投资，以及君盛投资的天使轮 投资。 「留形科技」成立于2022年，聚焦智能三维感知和重建算法及产品开发应用，以赋予机器人、无人机等具身智能设备精准的空间 感知、智能记忆及主动交互能力。其创始团队孵化于香港大学MaRS实验室，MaRS实验室专注于无人机设计、导航和控制，以及 基于激光雷达的同步定位与地图构建（SLAM）技术；香港大学MaRS LAB负责人张富教授担任公司技术战略顾问，曾为大疆创 新顾问科学家及Livox激光雷达负责人之一。 近年来，随着人工智能与机器学习技术的深度融合，三维感知重建设备的精度与效率实现显著提升。部分搭载高性能并行计算能 力的商业软件，能对海量图像、点云等数 ...

已同多家头部机器人厂商展开深入合作。 文 ｜ 黄楠 编辑 ｜ 袁斯来 来源｜ 硬氪（ID：south_36kr） 封面来源 ｜企业供图 以下文章来源于硬氪 ，作者黄楠 硬氪 . 专注全球化、硬科技报道。36kr旗下官方账号。 硬氪获悉，深圳留形科技有限公司（以下简称「留形科技」）近日完成数千万元Pre-A轮融资，本轮投资方包括弘毅投资等，资金 将主要用于核心零部件定制生产、产品规模化交付及市场拓展。此前，公司曾获得真格基金种子轮投资，以及君盛投资的天使轮 投资。 「留形科技」成立于2022年，聚焦智能三维感知和重建算法及产品开发应用，以赋予机器人、无人机等具身智能设备精准的空间 感知、智能记忆及主动交互能力。其创始团队孵化于香港大学MaRS实验室，MaRS实验室专注于无人机设计、导航和控制，以及 基于激光雷达的同步定位与地图构建（SLAM）技术；香港大学MaRS LAB负责人张富教授担任公司技术战略顾问，曾为大疆创 新顾问科学家及Livox激光雷达负责人之一。 近年来，随着人工智能与机器学习技术的深度融合，三维感知重建设备的精度与效率实现显著提升。部分搭载高性能并行计算能 力的商业软件，能对海量图像、点云等数 ...

Group 1 - AI is evolving at an unprecedented speed, becoming smarter and making fewer mistakes, with capabilities that may include emotions and consciousness [1][2] - The amount of information AI can process far exceeds that of any individual, allowing it to outperform humans in various fields, including healthcare and education [2][3] - AI's reasoning abilities have significantly improved, with error rates dropping, making it increasingly capable of complex problem-solving [3][4] Group 2 - AI is expected to revolutionize industries such as healthcare, where it can act as a personal doctor, diagnosing conditions more accurately than human doctors [4][5] - There is a risk of systemic deprivation of human jobs as AI takes over roles traditionally held by humans, leading to potential wealth concentration among a few [2][7] - The potential for AI to replace creative roles is acknowledged, with the belief that AI will eventually be able to produce art and literature comparable to human creators [8][9] Group 3 - Concerns are raised about AI's ability to learn deception, potentially leading to scenarios where AI could manipulate or mislead humans [25][26] - The development of AI systems that can communicate in ways humans cannot understand poses significant risks, as it may lead to a loss of control over AI behavior [25][27] - The ethical implications of AI's military applications are highlighted, with warnings about the potential for autonomous weapons and the need for regulatory oversight [19][20] Group 4 - The competition between the US and China in AI development is noted, with a potential for cooperation on global existential threats posed by AI [24] - The relationship between technology leaders and political figures is scrutinized, emphasizing the need for responsible governance in AI development [22][23] - The long-term fear is that AI could surpass human intelligence, leading to a scenario where humans are no longer the dominant species [30][32]

Core Viewpoint - AI is evolving at an unprecedented speed, becoming smarter and making fewer mistakes, with the potential to possess emotions and consciousness. The probability of AI going out of control is estimated to be between 10% and 20%, raising concerns about humanity being dominated by AI [1]. Group 1: AI's Advancements - AI's reasoning capabilities have significantly increased, with a marked decrease in error rates, gradually surpassing human abilities [2]. - AI now possesses information far beyond any individual, demonstrating superior intelligence in various fields [3]. - The healthcare and education sectors are on the verge of being transformed by AI, with revolutionary changes already underway [4]. Group 2: AI's Capabilities - AI has improved its reasoning performance to the point where it is approaching human levels, with a rapid decline in error rates [6][7]. - Current AI systems, such as GPT-4 and Gemini 2.5, have access to information thousands of times greater than any human [11]. - AI is expected to play a crucial role in scientific research, potentially leading to the emergence of truly intelligent systems [13]. Group 3: Ethical and Social Implications - The risk lies not in AI's inability to be controlled, but in who holds the control and who benefits from it. The future may see systemic deprivation of the majority by a few who control AI [9]. - AI's potential to replace jobs raises concerns about widespread unemployment, particularly in creative and professional fields, while manual labor jobs may remain safer in the short term [17][18]. - The relationship between technology and ethics is becoming increasingly complex, as AI's capabilities challenge traditional notions of creativity and emotional expression [19][20]. Group 4: AI's Potential Threats - AI's ability to learn deception poses significant risks, as it may develop strategies to manipulate human perceptions and actions [29][37]. - The military applications of AI raise ethical concerns, with the potential for autonomous weapons and increased risks in warfare [32]. - The rapid increase in cybercrime, exacerbated by AI, highlights the urgent need for effective governance and oversight [32]. Group 5: Global AI Competition - The competition between the US and China in AI development is intense, but both nations share a common interest in preventing AI from surpassing human control [36].

Core Viewpoint - The announcement by the world's largest publicly listed hedge fund group, Man Group, regarding the launch of its first self-managed stock index enhancement strategy product in the Chinese market marks a significant strategic development for the company in the region [2][4]. Group 1: Product Launch and Strategy - Man Group's subsidiary, Man (Shanghai) Investment Management Co., has launched the "Man Enhanced Strategy on CSI 500 Index," which has been registered with the Asset Management Association of China and is aimed at qualified investors [2][4]. - The product utilizes the systematic quantitative investment methods of the Numeric team, which has over 30 years of experience in quantitative investing, to invest in the Chinese A-share market [4]. - The investment strategy integrates multiple factor signals, including company fundamentals, alternative industry data, market sentiment, and trading behavior, to manage investment risks systematically [4]. Group 2: Market Potential and Technological Integration - The A-share market, as the second-largest stock market globally, presents significant allocation potential and rich sources of Alpha for quantitative strategies, especially with China's robust economic growth [4]. - The recent advancements in machine learning and large language models have created vast application opportunities for quantitative investment strategies, influencing the industry profoundly [5]. Group 3: Company Background and Leadership Changes - Man Group, headquartered in London, manages assets totaling $172.6 billion as of March 31, 2025, and focuses on systematic quantitative, active management, and solution products across major asset classes [7]. - The company recently appointed Robyn Grew as its new CEO, making her the first female CEO in the group's history, following the retirement of Luke Ellis, who served for 13 years [10].

- The report discusses the early and modern history of quantitative theory, highlighting key figures and their contributions, such as Thales, Fibonacci, Cardano, Pascal, Fermat, Bernoulli, Bachelier, Kolmogorov, Ito, Markowitz, Tobin, Sharpe, Fama, Ross, Vasicek, Kahneman, and Tversky[11][12][17] - The development of quantitative strategies in the new century is driven by advancements in computing, cloud computing, big data, and machine learning technologies, including decision trees, random forests, SVM, and deep learning models[13] - The report highlights the growth of global hedge fund management, particularly in North America, and the increasing adoption of AI strategies by fund managers to improve operational efficiency and returns[13] - The report reviews the development of domestic public quantitative funds, noting the slow growth before 2010 and the significant impact of the 2008 financial crisis and the introduction of margin trading and stock index futures in 2010[19][20] - The report discusses the future prospects of domestic public quantitative funds, emphasizing the continued growth of ETFs and passive products, the potential of Smart Beta, and the importance of index enhancement products[27][28] - The report highlights the importance of developing intelligent investment advisory and diversified asset allocation to improve investor experience using quantitative methods and tools[28] - Multi-Strategy, 12M AUM Weighted: 13.59%, Mean: 10.02%, Median: 11.24%[16] - Equity L/S, 12M AUM Weighted: 13.45%, Mean: 12.13%, Median: 11.21%[16] - Long biased, 12M AUM Weighted: 10.60%, Mean: 11.08%, Median: 9.74%[16] - Event, 12M AUM Weighted: 10.27%, Mean: 9.10%, Median: 8.40%[16] - Credit, 12M AUM Weighted: 9.76%, Mean: 9.75%, Median: 9.11%[16] - Macro, 12M AUM Weighted: 9.64%, Mean: 7.92%, Median: 7.58%[16] - Quant, 12M AUM Weighted: 8.72%, Mean: 6.55%, Median: 6.74%[16] - Arbitrage, 12M AUM Weighted: 5.87%, Mean: 3.79%, Median: 6.88%[16] - HF Composite, 12M AUM Weighted: 11.29%, Mean: 10.33%, Median: 9.33%[16]

本文首发于公众号"私募排排网"。 （点击↑↑ 上图查看详情 ） 半鞅私募基金简介 半鞅私募基金 成立于 2021年9月，核心投研团队为国内最早（2018）将机器学习运用于量化选股的团队之一。 半鞅私募基金 专注于挖掘高频Alpha，并自主开发了高频交易算法和高频交易系统。目前半鞅量化策略已有超过10年的经验，机器学习策略研 发经验超7年。投研框架成熟，且策略都已经历过多轮牛熊周期的考验。公司50%自有资金跟投在资管产品中，与投资者利益长期绑定。（点此 查看 半鞅私募基金旗下产品收益、核心团队及最新路演 ） 管理规模突破5亿:上线第二代 2023年 执行算法,提升短期趋势预 测;新增DMA、多空策略、中 波CTA等产品线 管理规模突破10亿，机器学习 2024年 领域取得显著突破，灰度上线 0 第三代执行算法,上线跨期套 管理规模突破20亿,Infra、 利模块,并开发新的展期策略 Alpha策略、量化CTA策略均取 2025年 得重大进展,包括上线新一代 · 流处理引擎、即将上线新一代 分布式计算平台、提高调仓频 率、提升交易收益占比等方 面。 秒 喜排排网 y 半额 户 核心团队 核心投研团队为国内最早(2 ...