Core Viewpoint - The article discusses the emergence of "Sovereign AI," a strategic effort by nations and organizations to develop, deploy, and govern AI capabilities independently, minimizing external dependencies. This reflects a collective anxiety about digital autonomy and control over one's data and future [1]. Group 1: Strategic Consensus - The pursuit of AI sovereignty has become a global strategic consensus, with 79% of respondents valuing the development of AI capabilities that reduce external dependencies [3][4]. - This consensus transcends geographical boundaries, with 86% in North America, 83% in Europe, and 79% in the Asia-Pacific region recognizing its strategic relevance [6]. Group 2: Key Drivers - Four core drivers propel the global movement towards Sovereign AI: 1. Data Sovereignty and Control (72%): The desire to control data as a strategic asset to avoid "digital colonialism" [8]. 2. National Security (69%): The control of AI systems is crucial for safeguarding national security, especially concerning critical infrastructure [9]. 3. Economic Competitiveness (48%): Sovereign AI is seen as essential for building domestic innovation ecosystems and enhancing global competitiveness [10]. 4. Cultural Fit and Regulatory Compliance (31% and 44%): The need for AI to reflect local culture and comply with regulations like GDPR is significant [11]. Group 3: Paradox of Implementation - The article highlights a paradox in achieving Sovereign AI, where the need for independence conflicts with the necessity of global collaboration. A staggering 94% of respondents believe global cooperation is essential for realizing Sovereign AI [14][16]. - Open source is proposed as a solution to this paradox, providing transparency, flexibility, and security, which are crucial for building trust and control in AI systems [17][18]. Group 4: Future Pathways - The report identifies significant challenges on the path to open-source Sovereign AI, including data quality and availability (44%), technical expertise shortages (35%), and security vulnerabilities (34%) [23]. - Different regions face unique challenges, with the U.S. focusing on data quality, Europe on compliance, and Asia-Pacific on security vulnerabilities and skill shortages [26]. Group 5: Governance Models - The future governance of AI is expected to be a decentralized model involving governments, open-source communities, academia, and industry, rather than a top-down approach [30][31].

Investment Rating - The report does not explicitly provide an investment rating for the computing power scheduling platform industry. Core Insights - The rapid development of artificial intelligence technology has led to an exponential increase in global demand for computing power, necessitating computing power scheduling for resource integration and optimization across regions and platforms [2]. Summary by Sections Overview of the Computing Power Scheduling Industry - Computing power is defined as the ability of computer devices or data centers to process information, categorized into general computing power, intelligent computing power, and supercomputing power [15][18]. - China's computing power scale has grown rapidly, reaching 280 EFLOPS by 2024, with intelligent computing power accounting for 32% [20][23]. Challenges in Heterogeneous Computing Power Scheduling - Heterogeneous computing power scheduling faces multiple core challenges, including increased scheduling complexity due to resource heterogeneity and software environment fragmentation, high migration costs for cross-architecture tasks, and a lack of unified scheduling standards leading to resource mismatch and low utilization [4][43]. Major Domestic Computing Power Scheduling Platforms - National-level computing power scheduling platforms are primarily government-led or constructed by major operators, emphasizing cross-regional collaboration and market-oriented transactions [5][48]. - Provincial platforms cover key regions like the Yangtze River Delta and Chengdu-Chongqing, while municipal platforms focus on local AI and smart manufacturing scenarios [48]. Mainstream Open Source Computing Power Scheduling Technologies - Domestic computing power scheduling platforms are often built on open-source technologies, with openFuyao emerging as a versatile scheduling platform with advantages in domestic adaptation, while Kubernetes and Slurm have strong foundations in cloud-native and HPC fields [6][51].

Group 1 - The core viewpoint emphasizes that cloud-native and microservices architecture are central to enterprise technology upgrades amid digital transformation [2][3] - The emergence of the "Mago Cloud Native Microservices Governance Sprint Class" reflects a deep transformation in technical education and reveals a new interaction logic among technology, economy, and talent market [2] - Traditional IT training is undergoing a paradigm shift, focusing on reshaping architectural thinking rather than just skill training [2] Group 2 - The course integrates tools like Kubernetes and Istio with microservices governance methodologies, enabling students to transition from merely coding to mastering distributed systems [2] - The curriculum addresses industry pain points, with governance capabilities such as service discovery and circuit breaking becoming essential as companies migrate from monolithic to microservices architecture [3] - The economic effect shows that the cloud computing industry has surpassed a trillion in scale, with programmers skilled in cloud-native technologies enjoying a 40% salary premium [3] Group 3 - The course's continuous update mechanism reflects the necessity of lifelong learning for technical professionals, especially in the face of AI's impact on traditional programming roles [3] - The transformation to cloud-native skills has allowed professionals to move from passive roles to core architecture teams, enhancing their job security [3] - The upgrade in technical capabilities driven by high-quality courses signifies a reconstruction of production relationships in the digital age, contributing to a more efficient and resilient technological ecosystem [3]

Core Technology & Product Offering - Daily 公司提供实时音视频和 AI 的全球基础设施，并推出开源、供应商中立的项目 Pipecat，旨在帮助开发者构建可靠、高性能的语音 AI 代理 [2][3] - Pipecat 框架包含原生电话支持，可与 Twilio 和 Pivo 等多个电话提供商即插即用，还包括完全开源的音频智能转向模型 [12][13] - Pipecat Cloud 是首个开源语音 AI 云，旨在托管专为语音 AI 问题设计的代码，支持 60 多种模型和服务 [14][15] - Daily 推出 Pipecat Cloud，作为 Docker 和 Kubernetes 的轻量级封装，专门为语音 AI 优化，解决快速启动、自动缩放和实时性能等问题 [29] Voice AI Agent Development & Challenges - 构建语音代理需要考虑代码编写、代码部署和用户连接三个方面，用户对语音 AI 的期望很高，要求 AI 能够理解、智能、会话且听起来自然 [5][6] - 语音 AI 代理需要快速响应，目标是 800 毫秒的语音到语音响应时间，同时需要准确判断何时响应 [7][8] - 开发者使用 Pipecat 等框架，以避免编写turn detection（转弯检测）、中断处理和上下文管理等复杂代码，从而专注于业务逻辑和用户体验 [10] - 语音 AI 面临长会话、低延迟网络协议和自动缩放等独特挑战，冷启动时间至关重要 [25][26][30] - 语音 AI 的主要挑战包括：背景噪音会触发不必要的LLM中断，以及代理的非确定性 [38][40] Model & Service Ecosystem - Pipecat 支持多种模型和服务，包括 OpenAI 的音频模型和 Gemini 的多模态实时 API，用于会话流程和游戏互动 [15][19][22] - 行业正在探索 Moshi 和 Sesame 等下一代研究模型，这些模型具有持续双向流架构，但尚未完全准备好用于生产 [49][56] - Gemini 在原生音频输入模式下表现良好，且定价具有竞争力，但模型在音频模式下的可靠性低于文本模式 [61][53] - Ultravox 是一个基于 Llama 3 7B 主干的语音合成模型，如果 Llama 3 70B 满足需求，那么 Ultravox 是一个不错的选择 [57][58] Deployment & Infrastructure - Daily 公司在全球范围内提供端点，通过 AWS 或 OCI 骨干网路由，以优化延迟并满足数据隐私要求 [47] - 针对澳大利亚等地理位置较远的用户，建议将服务部署在靠近推理服务器的位置，或者在本地运行开放权重模型 [42][44] - 语音到语音模型的主要优势在于，它们可以在转录步骤中保留信息，例如混合语言，但音频数据量不足可能会导致问题 [63][67]

Core Insights - The article discusses the establishment of the Agent2Agent (A2A) project by the Linux Foundation in collaboration with major tech companies like AWS, Google, and Microsoft, aimed at creating an open standard for communication between AI agents [1][3][7] - A2A is positioned as a higher-level protocol compared to the Model Context Protocol (MCP), facilitating seamless interaction among multiple AI agents, while MCP focuses on integrating large models with external tools [6][7][11] - The article highlights the importance of these protocols in enhancing the reliability and functionality of AI systems, particularly in complex workflows involving multiple AI agents [14][15][18] Summary by Sections A2A Project Announcement - The A2A project was announced at the North America Open Source Summit on June 23, with initial contributions from Google, including the A2A protocol specification and related SDKs [1] - The A2A protocol aims to address the "island" problem of AI by enabling communication and collaboration between different AI systems [1] Comparison with MCP - MCP has rapidly expanded, growing from 500 servers in February to over 4000 servers currently, indicating its swift adoption [4] - A2A operates at a higher level than MCP, focusing on inter-agent communication, while MCP standardizes communication between large models and external tools [6][7] Developer Perspectives - Developers express uncertainty about how A2A and MCP will coexist, with some suggesting that A2A needs to demonstrate unique capabilities to stand out [11] - A2A's HTTP-based communication model may offer easier integration compared to MCP, which has been noted for its complexity [11][12] Protocol Necessity and ROI - The necessity of adopting these protocols is questioned, with some industry leaders suggesting that they should only be used when genuinely needed [13] - The article emphasizes the challenges in measuring ROI for AI applications, highlighting that only about 5% of generative AI projects have turned into profitable products [18] Security and Monitoring Concerns - There are concerns regarding the security and complexity of both protocols, particularly in terms of identity verification and authorization [17] - The monitoring and evaluation mechanisms for agent-driven systems are still in early stages, indicating a need for further development in this area [17]

Core Insights - China and Hong Kong have emerged as one of the earliest and strongest ecosystems in the cloud-native field, with a total of 1.0686 million open-source contributions, ranking second globally, including 327,400 contributions to the Kubernetes project [1] - The KubeCon+CloudNativeCon China 2025 summit, held in Hong Kong, signals the growing importance of the region in the cloud-native landscape [1] - The cloud-native concept, defined by CNCF, facilitates the construction and operation of scalable applications in dynamic environments such as public, private, and hybrid clouds [1] Group 1: Ecosystem Growth - CNCF's 2024 annual report indicates over 140 new members joined last year, bringing the total to over 200 projects and 728 members, with more than 270,000 contributors from 189 countries [4] - Key open-source projects from China, such as Volcano, Dragonfly, KubeEdge, and OpenYurt, demonstrate significant capabilities in edge computing, container scheduling, and distributed processing [2] Group 2: Industry Applications - Major cloud service providers like Tencent Cloud, Huawei, Alibaba Cloud, and Baidu Intelligent Cloud are members of the ecosystem, contributing to advancements in distributed consensus mechanisms within Kubernetes [5] - Hong Kong's financial institutions are core adopters of cloud computing technology, with platforms like the new IPO settlement platform FINI and HKEX Synapse enhancing digital transaction processes [5] Group 3: AI Integration - Cloud-native computing technologies are expected to bring systemic innovations to the AI industry, with local deployments by major companies supporting digitalization needs in Hong Kong [6] - The Hong Kong government views cloud computing and cloud-native technologies as key foundations for smart city development, promoting their application in e-government, smart transportation, and healthcare [8] Group 4: Open Source Impact - The economic value of open source is highlighted, with a study indicating that the cost of acquiring all necessary open-source software for technology creation could reach $9 trillion [6] - The success of open-source projects relies on user participation and a structured approach that is friendly to new contributors [8]

Investment Rating - The report indicates a positive investment outlook for the computing power scheduling platform industry, driven by the rapid growth in AI model applications and the increasing demand for high-performance computing resources [10][14]. Core Insights - The demand for intelligent computing power in China is experiencing unprecedented growth, particularly in large model applications, which account for nearly 60% of the demand, highlighting the significant potential of the future computing power market [14][30]. - The profitability of computing power scheduling centers heavily relies on government subsidies, which are designed to ensure local resource utilization and risk control. Successful engagement with government decision-making is crucial for companies to secure these subsidies [22][30]. - The core value of computing power scheduling platforms lies in their ability to efficiently integrate and schedule heterogeneous computing resources, significantly improving resource utilization and reducing costs for users [17][20]. Summary by Sections Computing Power Scheduling Platform's Scaling Path - The rapid growth of intelligent computing demand in China is driven by AI large models, leading to accelerated development of large-scale, high-performance computing centers [11][10]. - The profitability of computing power scheduling centers is highly dependent on government subsidies, which aim to ensure local resource utilization and risk management [22][30]. - Platforms must possess efficient integration and scheduling capabilities for heterogeneous computing resources to achieve low-cost, scalable, and marketable monetization [31][39]. Value of Computing Power Scheduling Platforms - The platforms enhance resource utilization, lower user costs, and simplify management processes, providing efficient and convenient computing power services [17][20]. - The core technologies required include resource virtualization, fine-grained slicing, real-time monitoring, and tidal scheduling, which enable low-cost and efficient utilization of resources [31][38]. Government Subsidies as Core Source - Government subsidies are essential for the profitability of computing power scheduling centers, with a structured mechanism to ensure local resource utilization and risk control [22][30]. - Companies must strategically engage early in government decision-making to influence standards and secure contracts [23][30]. Technical Features of Quality Computing Power Scheduling Platforms - Platforms need to have capabilities for fine-grained resource slicing, heterogeneous compatibility, cross-regional scheduling, real-time monitoring, and dynamic scheduling to achieve efficient resource reuse and low-cost monetization [31][38]. Core of Scaling Computing Power Centers - The core of monetizing the value of computing power platforms lies in a large and diverse customer base, which determines profitability speed and pricing potential [39][42]. - A diverse customer base allows for tiered pricing strategies to maximize revenue, while partnerships can help focus on high-margin computing power sales [42][39].

Core Points - The article discusses the architecture and technology choices for social applications, emphasizing the importance of selecting the right frameworks and services for development [5][8][9]. Group 1: Frontend Development - The frontend of a social app consists of mobile (iOS/Android) and web applications, utilizing frameworks like React.js, Vue.js, and Angular for single-page applications [3][5]. - Mobile app development can be native (using Swift for iOS and Kotlin for Android) or cross-platform (using React Native, Flutter, uni-app, or Taro), each with its own advantages and disadvantages [6][8]. Group 2: Backend Development - The backend handles business logic, data storage, user authentication, and API interfaces, with popular frameworks including Spring Boot for Java, Django for Python, and Express.js for Node.js [9]. - Java is noted for its high performance and stability, making it suitable for large-scale applications, while Python offers rapid development capabilities for smaller projects [9]. Group 3: Database and Storage Solutions - Relational databases like MySQL and PostgreSQL are commonly used for structured data, while NoSQL databases like MongoDB and Redis are preferred for unstructured data and high-speed access [9]. - Object storage services from providers like Alibaba Cloud and Tencent Cloud are essential for managing user-generated content such as images and videos [9]. Group 4: Cloud Services and Compliance - For the Chinese market, compliance with local regulations, including ICP filing and app registration, is crucial, along with the selection of domestic cloud service providers like Alibaba Cloud and Tencent Cloud [8]. - The article highlights the importance of integrating third-party SDKs for functionalities like instant messaging and content moderation, with a focus on local providers [8][9]. Group 5: Development Tools and Technologies - The use of message queues (e.g., Kafka, RabbitMQ) and search engines (e.g., Elasticsearch) is recommended for system decoupling and enhancing user experience through personalized content [9]. - Containerization technologies like Docker and Kubernetes are suggested for efficient application deployment and management [9].

Core Insights - Cloud computing is one of the fastest-growing sectors in technology, characterized by the delivery of computing services over the internet, allowing organizations to scale resources efficiently [1][3] - The sector benefits from economies of scale, where profitability growth can significantly exceed revenue growth once fixed costs are covered [2] - The rise of artificial intelligence (AI) has accelerated growth in cloud computing as organizations utilize cloud services to develop and run AI models and applications [3] Company Summaries Amazon - Amazon is the largest cloud computing service provider globally, holding nearly a 30% market share, with its Amazon Web Services (AWS) segment being the most profitable and fastest-growing [6] - AWS revenue increased by 17% year-over-year to $29.3 billion, while operating income rose by 22% to $11.5 billion [6] - Key growth drivers for AWS include its Bedrock and SageMaker solutions, which allow customers to customize AI models and build their own from scratch [7][8] Microsoft - Microsoft Azure has been gaining market share, with revenue growth of 30% or more for the past seven quarters, reaching a market share of around 22% [9] - The partnership with OpenAI has enhanced Azure's offerings, allowing customers to integrate leading AI models into their applications [10] - Microsoft is diversifying its AI portfolio by hosting models from xAI and hiring talent from DeepMind to develop its own AI models [11] Alphabet - Alphabet's Google Cloud, with about a 12% market share, has reached a profitability inflection point, with revenue climbing 28% year-over-year to $12.3 billion and operating income surging 142% to $2.2 billion [12][13] - Google Cloud's competitive edge comes from its Vertex AI platform, analytics tools like BigQuery, and leadership in Kubernetes [14] - Alphabet has developed advanced AI models like Gemini and custom AI chips to enhance its cloud services, despite concerns about AI's impact on its search business [15][16]

Thank you. Um I'm Jen from SK Terracon from South Korea and uh today I'd like to give a talk about the storage virtualization and optimization for GPU as a service that what we are trying to do recently. And uh first of all I think you are not quite familiar what is SQL.So I would like to give a brief like introduction to the SK talon. JSK is a novel metal player in South Korea and we made the most of the market penetrations in South Korea and recently we are trying to transform from the NNO to the AI compa ...