9月24日，2025云栖大会在杭州云栖小镇国际会展中心拉开帷幕。作为与阿里合力打造人工智能产品、 方案和服务的合作伙伴，联想集团有限公司（以下简称"联想"）联合"全栈AI"再次亮相云栖大会，围 绕"混合式AI让AI成为创新生产力"主题，全面展示了涵盖AI终端、AI基础设施再到AI解决方案及服务 的全场景算力产品、方案和服务。 其中AI基础设施方面，聚焦智算、科学计算与通算三大计算场景，联想展出了覆盖全场景的领先算力 基础设施产品、解决方案，以及全球领先的相变浸没式液冷技术。而围绕AI终端和AI解决方案及服 务，天禧个人超级智能体、创新AI终端、智算中心、智能体矩阵等也圈粉无数。 本届云栖大会以"云智一体，碳硅共生"为主题，全景式呈现了云计算以及相关领域的产业趋势、技术创 新、关键产品和前沿应用。 算力全面跃升筑牢发展底座 伴随AI普惠时代的到来，AI正在成为新时代创新生产力的"核心引擎"。在AI领域，联想发力很早。早 在2017年，联想就前瞻布局人工智能，提出"联想智慧中国"的愿景。通过八年的创新与探索，联想基 于"端-边-云-网-智"新IT架构，着手构建了"全栈AI"战略布局，积累了丰厚的产品、方案和服务等 ...

Core Insights - The announcement of a $455 billion unfulfilled revenue obligation (RPO) by a North American tech giant has significantly boosted the global computing power sector, highlighting strong demand for AI infrastructure and a supply-demand imbalance in computing power [1] - The North American contract signifies a strategic turning point for the global computing industry, elevating computing power to a strategic asset level comparable to oil and chips [6] Industry Overview - IDC predicts that the global AI market will exceed $800 billion by 2028, with a compound annual growth rate (CAGR) of over 32% [2] - Morgan Stanley forecasts global data center spending to approach $3 trillion by 2029 [2] - The China Academy of Telecommunications Research estimates that AI could contribute over 11 trillion yuan to China's GDP by 2035, potentially driving a tenfold to hundredfold increase in computing power demand [2] Policy Developments - Recent policies aim to implement "AI+" actions across various sectors, promoting deep integration of AI with economic and social fields [4] - The Ministry of Industry and Information Technology (MIIT) plans to guide the construction of computing power facilities to enhance resource supply quality [4] Demand Dynamics - The computing power arms race is intensifying, with new technologies like DeepSeek enabling low-cost industrialization of large models, stimulating new application scenarios and reducing inference barriers [5] - IDC forecasts an 18% growth in the global infrastructure market by 2025, with AI servers expected to reach a market size of $147.2 billion [5] Company Positioning - Lenovo is strategically positioned to capitalize on the AI wave, leveraging its "computing gene" to secure a competitive edge in the market [6] - Lenovo's hybrid AI framework integrates various models to address traditional AI deployment challenges, enhancing efficiency and sustainability [7] Product Innovations - Lenovo has launched the next-generation ThinkSystem V4 servers and ThinkEdge SE100 AI inference servers, enhancing its competitiveness in the AI server market [8] - The company's liquid cooling solutions demonstrate superior thermal management and energy efficiency, aligning with industry trends towards energy conservation [8] Financial Performance - Lenovo's AI infrastructure business has seen explosive growth, with a 155% year-on-year revenue increase and strong order reserves [10] - The company's liquid cooling technology has achieved nearly 30% revenue growth, contributing to its overall financial success [10] - Analysts maintain a positive outlook on Lenovo, with firms like Huatai Securities and Guotai Junan issuing "buy" and "overweight" ratings, respectively, due to the rapid growth in global AI infrastructure demand [10][11]

Group 1 - The seminar themed "Heterogeneous Fusion Innovation Future, Intelligent Computing Empowering Marine Technology" was held in Qingdao, focusing on the innovative applications of GPU+CPU hybrid architecture in deep-sea exploration and marine carbon sink monitoring [1] - The national-level platform for marine technology, Laoshan Laboratory, is integrating marine technology with heterogeneous computing to provide strong strategic technological support for building a marine power [2] - The rapid development of large model technology is providing a solid foundation for marine artificial intelligence, with Qingdao Guoshi Group actively promoting the transformation and industrialization of technological achievements in the marine technology sector [2] Group 2 - The integration of computing power, models, and applications is reshaping the marine economy, with the demand for computing power driving growth in high-performance computing and intelligent computing markets [3] - Lenovo's AI infrastructure strategy includes a heterogeneous computing platform designed to meet the computing needs of large model pre-training and traditional model inference, featuring five key technologies [3][4] - Intel is continuously updating and enhancing its product portfolio to support diverse computing power needs across various industries in the context of rapid AI development [4]

Core Viewpoint - The seminar focused on the innovative applications of GPU+CPU hybrid architecture in deep-sea exploration and marine carbon sink monitoring, highlighting the collaboration between the heterogeneous computing industry and various partners, including Intel [1][5][6]. Group 1: Event Overview - The seminar titled "Innovating the Future of Heterogeneous Integration, Empowering Marine Technology with Intelligent Computing" was held in Qingdao, featuring discussions on the innovative applications of hybrid computing architectures [1]. - Key attendees included members from the Chinese Academy of Sciences, the Marine Technology Industry Alliance, and various technology companies, who engaged in in-depth discussions and showcased technological advancements [3][5]. Group 2: Technological Insights - The integration of AI, high-performance computing, and cloud computing is enhancing the understanding and predictive accuracy of marine environmental changes, contributing to sustainable development goals [5]. - The rapid development of large model technology is providing a solid foundation for marine artificial intelligence, enabling better understanding of climate change impacts on marine ecosystems [9]. Group 3: Strategic Initiatives - The Laoshan Laboratory is focusing on the integration of marine technology and heterogeneous computing to support the construction of a strong marine nation, emphasizing the importance of high-performance computing systems [6]. - Qingdao Guoshi Group aims to leverage advancements in algorithms, computing power, and data accumulation to accurately simulate marine ecosystems and address data processing challenges in various marine technology fields [9]. Group 4: Industry Collaboration - Lenovo is advancing its "One Horizontal and Five Verticals" strategy to provide comprehensive infrastructure solutions, including a heterogeneous computing platform tailored for large model training and inference [12][13]. - Intel is committed to developing HPC & AI integrated computing technologies to support diverse industry needs and enhance AI technology deployment [15][17].

Core Insights - The domestic computing power ecosystem is evolving through collaboration across various sectors, from chips to servers and intelligent computing clusters, aiming for higher efficiency and application deployment [1][2][3] - Companies like DeepSeek are leading the charge in integrating domestic chips into practical applications, enhancing computational efficiency while reducing storage and data transmission costs [4][5] - The launch of the OISA 2.0 protocol during the conference marks a significant step in building a collaborative platform for GPU interconnectivity, supporting the scale-up of intelligent computing clusters [5][6] Industry Developments - The collaboration among domestic operators, internet companies, chip manufacturers, and research institutions is crucial for establishing a cohesive computing power industry chain [3][5] - The OISA 2.0 protocol supports up to 1024 AI chips with bandwidth exceeding TB/s and latency reduced to hundreds of nanoseconds, enhancing the performance of intelligent computing clusters [5] - The introduction of the GSE technology system by China Mobile aims to optimize the scale-out route for intelligent computing centers, focusing on high-capacity networking capabilities [6] Technological Innovations - The industry is addressing the challenges of heterogeneous computing by developing unified platforms that enhance ecosystem synergy [6][8] - The integration of high-performance computing and intelligent computing requires a deep restructuring of hardware architecture, emphasizing the need for collaborative innovation across various layers [8] - The focus on liquid cooling technologies is increasing, with cold plate liquid cooling systems being highlighted for their efficiency in high-density deployments [11][12] Market Trends - The demand for intelligent computing centers is rising, but challenges remain in infrastructure planning, model development efficiency, and deep integration of industrial applications [9][10] - The report emphasizes the need for a comprehensive standard system covering construction, development, and application processes in intelligent computing services [9] - Companies are actively developing full-stack solutions to meet diverse computing power demands across various industries, including education, energy, and healthcare [10] Future Directions - The industry is moving towards a collaborative ecosystem that promotes open protocols and integrated solutions, driving technological advancements in the domestic computing power sector [13] - The focus on energy efficiency and cost reduction through innovative cooling solutions is expected to play a critical role in the future of data center construction [11][13]

Core Insights - Lenovo Group, in collaboration with the Heterogeneous Computing Industry Alliance, released the "2025 AI Empowerment White Paper" at the 2025 China Computing Power Conference, showcasing its deep technological foundation in AI infrastructure and insights into customer needs [1][2] - This white paper is the second released since the establishment of the Heterogeneous Computing Industry Alliance last year, analyzing technological evolution trends and providing actionable solutions for intelligent transformation across various industries [1][2] Summary by Sections - **AI Core Technologies**: The white paper highlights three innovative AI core technologies: reinforcement learning framework optimization, large model training and inference technology, and innovative long-term storage technology for industrial quality inspection in manufacturing [1] - **Lenovo's Heterogeneous Computing Platform**: The large model training and inference technology refers to Lenovo's Wanquan Heterogeneous Computing Platform, which has been upgraded to version 3.0. It includes nine core capabilities covering the entire process of pre-training, post-training, and inference, providing comprehensive AI computing support for enterprise users [2] - **Practical Applications**: The platform has been validated in customer scenarios, such as optimizing the DeepSeek large model for a fund company, resulting in training costs being only 10% of similar models, inference costs reduced to 2%-3%, and an AI substitution rate of 30%-50%. Additionally, Lenovo collaborated with Geely to create a benchmark intelligent computing cluster for the manufacturing industry, achieving cost optimization and smooth business upgrades in hybrid computing scenarios [2]

Core Insights - Lenovo is actively embracing new AI opportunities, having laid a comprehensive foundation for artificial intelligence since 2017, which includes AI terminals, infrastructure, and solutions [2][3] - The company is collaborating with authoritative institutions to establish the industry's first high-performance training and inference service standards, and has released the "2025 AI Empowerment White Paper" [2][4] Group 1: AI Development Strategy - Lenovo's Vice President, Chen Zhenkuan, emphasized that computing power, models, and applications are the three main lines driving AI development, which are interlinked to propel the AI trend [2] - The company is leading a transformative shift from IT services to AI services through its "One Engine, Three Arrows" strategy, enhancing its solution and service capabilities [2][3] Group 2: AI Infrastructure - Lenovo is constructing a diverse AI computing power infrastructure, featuring a product portfolio that includes large model training, inference, and edge computing servers, along with a "1+3+N" long-term development framework [3] - The company has upgraded its heterogeneous intelligent computing platform to version 3.0, introducing four new innovative technologies to support model development trends [3] Group 3: Industry Collaboration - Lenovo, in partnership with the Heterogeneous Intelligent Computing Industry Ecosystem Alliance and various ecological partners, has released the latest "2025 AI Empowerment White Paper," which sets new industry benchmarks for AI computing technology and solutions [4] - Chen Zhenkuan expressed the need for deep integration among computing power providers, model developers, and application enablers to promote intelligent transformation and achieve inclusive computing power [4]

Core Insights - Lenovo's Vice President Chen Zhenkuan emphasized the integration of computing power, models, and applications as the three main lines driving AI development, highlighting Lenovo's efforts in building diverse AI computing infrastructure [2] - Lenovo is collaborating with authoritative institutions to establish the industry's first high-performance training and inference service standards [2] - The company is leading the third generational upgrade of the PC industry through an "all-in-one multi-end" approach in AI terminals and transitioning from IT services to AI services with comprehensive upgrades in solutions and service capabilities [2] Computing Power - Lenovo offers a comprehensive product portfolio for AI full-scene server services, including large model training, inference, integrated training and inference, and edge inference [2] - The long-term development framework is structured as "1+3+N," which consists of one system architecture, three CPU platforms, and compatibility with multiple GPU manufacturers [2] Storage and Network Solutions - In the storage domain, Lenovo provides core storage solutions that are large-capacity, high-density, and energy-efficient, along with the latest all-flash software-defined storage [2] - The AI RoCE intelligent computing network solution has been launched to support high-speed, low-latency network requirements for AI scenarios [2] Heterogeneous Computing Platform - Lenovo's Wanquan Heterogeneous Intelligent Computing Platform has been upgraded to version 3.0, introducing four new innovative technologies to support model development trends [3] - The platform has demonstrated practical validation in customer scenarios, achieving a model floating-point computation utilization rate increase from 30% to 60% in national-level high-quality AI cluster scenarios [3] - The DeepSeek R1 model has achieved an extreme throughput of over 12,000 tokens per second in enterprise AI infrastructure scenarios [3] Industry Collaboration - Lenovo has partnered with the Heterogeneous Intelligent Computing Industry Ecosystem Alliance and various ecosystem partners to release the latest "2025 AI Solutions Empowerment White Paper," which outlines insights on industry trends and technological innovations [3]

Group 1 - The core message of the article highlights Lenovo's commitment to advancing AI technology through the development of a comprehensive AI stack, which includes AI terminals, infrastructure, and solutions [1][3] - Lenovo is collaborating with authoritative institutions to establish the industry's first high-performance training and inference service standards, as announced during the 2025 China Computing Power Conference [1][3] - The company has upgraded its heterogeneous computing platform to version 3.0, introducing four new innovative technologies to support the evolving trends in model development [3][4] Group 2 - Lenovo's AI infrastructure includes a product portfolio that supports various AI scenarios, such as large model training and edge inference, structured under a "1+3+N" long-term development framework [3] - The company has achieved significant improvements in AI model training efficiency, increasing the floating-point computation utilization rate from 30% to 60% in collaboration with the "East Data West Computing" initiative [4] - The newly released "2025 AI Empowerment White Paper" outlines Lenovo's insights on industry trends and technological innovations, developed in partnership with the heterogeneous computing industry ecosystem [4]

Core Insights - The WEC2025 World Computing Power Expo successfully took place in Ordos, showcasing a record 71 exhibitors and emphasizing the integration of computing power technology with green energy [1][2][12] - The expo focused on six key areas: artificial intelligence, computing infrastructure, computing services, education and research, urban cultural exchange, and various display activities [2] - The event highlighted the development of the computing power industry in Ordos, with a comprehensive exhibition area featuring 14 companies from the computing power supply chain [4] Industry Overview - The WEC2025 presented a panoramic view of the computing power industry, covering three core areas, 12 sub-industries, and 78 vertical tracks, establishing a complete ecosystem from computing foundation to application scenarios [3] - The expo featured significant advancements in liquid cooling technology, which is becoming mainstream due to its high efficiency and reliability, with notable products like the Nebula800 platform being showcased [6][7] - The event also highlighted the importance of AI and cloud computing, with companies like Beijing Parallel Technology demonstrating their parallel computing network architecture and services for major model enterprises [8] Technological Innovations - The expo showcased the first standardized ultra-intelligent integrated computing platform, Nebula800, which addresses complex computing tasks in various fields [6][7] - High-throughput Ethernet (ETH+) technology was prominently featured, with several key products launched, including a 400G smart network card and a 25.6T switch chip [9] - Lenovo presented its comprehensive solution based on the heterogeneous intelligent computing platform, emphasizing its strategic approach to scientific and intelligent computing [9] Networking and Collaboration - The expo served as a vital platform for industry exchange, connecting research breakthroughs with practical applications and facilitating collaboration among enterprises, research institutions, and government bodies [12] - The event is expected to drive the evolution of the computing power ecosystem, focusing on technological integration, regional collaboration, and enhancing public services through computing power [12]