Investment Rating - The report maintains a "Positive" outlook on the industry [5] Core Viewpoints - The CXL solution optimizes AI storage architecture, and leading manufacturers are expected to accelerate its application [3][10] - The demand for memory capacity in AI inference is increasing, and the current server memory expansion is limited by the number of slots and the capacity of individual memory modules [10][19] - CXL memory pooling solutions can significantly enhance storage system efficiency and reshape the hardware composition in AI computing facilities [8][10] Summary by Sections 1. CXL Solutions Optimize Storage Efficiency and Adapt to AI Inference Needs - CXL solutions help expand memory capacity and optimize storage architecture, addressing limitations in existing AI computing facilities [19] - CXL memory pooling can achieve unified addressing and scheduling of memory resources across CPUs, GPUs, and other accelerators, supporting larger-scale and higher-concurrency model training and inference tasks [21][24] - Innovations in CXL solutions are ongoing, further adapting to AI inference requirements [32][38] 2. CXL-Related Hardware and Software Are Gradually Improving, with Leading Manufacturers Accelerating Application - CXL specifications are continuously upgraded, with transmission rates increasing from 32 GT/s to 128 GT/s by 2025 [49] - Major manufacturers like NVIDIA and Alibaba are actively advancing CXL solutions, with NVIDIA acquiring Enfabrica to enhance its CXL technology ecosystem [58][65] 3. CXL Solution Penetration Rate Is Expected to Increase, Opening Growth Space for the Industry Chain - The penetration rate of CXL technology in server DRAM is projected to grow from nearly zero in 2024 to about 15% by 2030 [70] - By 2026, it is expected that 68% of servers will support CXL functionality, indicating a significant shift towards CXL-capable servers [72] 4. CXL Applications Are Expected to Accelerate, with Related Companies Deeply Benefiting - Companies such as 澜起科技 (Lanke Technology), 聚辰股份 (Jucheng Co.), and 江波龙 (Jiangbolong) are positioned to benefit from the acceleration of CXL applications [13][75] - 澜起科技 reported a revenue of 5.46 billion yuan in 2025, a 50% year-on-year increase, and a net profit of 2.24 billion yuan, up 58% [76][80] - 聚辰股份 achieved a revenue of 1.22 billion yuan and a net profit of 360 million yuan in 2025, marking historical highs [85] - 江波龙 launched its CXL 2.0 memory expansion module, contributing to its revenue growth of 72% in 2024 [91][93]

Investment Rating - The report maintains a "Positive" investment rating for the electronic industry, indicating an expectation of returns exceeding the market benchmark by more than 5% [5]. Core Insights - The industry is focusing on CXL technology, with Nvidia expected to drive advancements. Key investment targets include companies such as Lanqi Technology, Jucheng Co., Jiangbolong, and Baiwei Storage [3][9]. - The CXL memory pooling solution is anticipated to optimize storage system efficiency and reshape future AI storage architectures. This solution allows for unified addressing and scheduling of memory resources across CPUs, GPUs, and other accelerators, supporting larger-scale and higher-concurrency AI model training and inference tasks [8]. - The CXL memory pooling scheme is gradually being perfected, with leading manufacturers accelerating their applications. The CXL 4.0 specification, set to be released in November 2025, will double the data rate to 128 GT/s compared to CXL 3.0 [8]. - Continuous innovation in CXL memory pooling solutions is expected to further adapt to AI inference demands, enhancing model inference efficiency and addressing existing storage architecture issues [8]. Summary by Sections Investment Recommendations and Targets - The report emphasizes the importance of CXL technology in the industry, highlighting Nvidia's role in its advancement. Recommended investment targets include Lanqi Technology (688008, Buy), Jucheng Co. (688123, Not Rated), Jiangbolong (301308, Not Rated), and Baiwei Storage (688525, Not Rated) [3][9]. Industry Focus - The report discusses the growing demand for AI, which is driving a persistent shortage in NAND storage. The need for AI computing power is expected to continue influencing storage shortages [7]. - The CXL memory pooling solution is seen as a critical innovation that can alleviate current bottlenecks in memory resources, thereby enhancing the performance of AI applications [8].

Investment Rating - The industry investment rating is "Positive" (maintained) [1] Core Insights - The CXL protocol ecosystem is continuously improving, with major CPU manufacturers like Intel and AMD releasing CPUs that support CXL 2.0, enhancing compatibility and performance [5][11] - The demand for memory bandwidth and capacity is increasing in the AI era, leading to a need for solutions that address the "compute-storage imbalance" [12][13] - CXL technology allows for the construction of memory pools, enabling the decoupling of storage and compute resources, which is essential for optimizing resource utilization across devices [14][15] Summary by Sections CXL Protocol Ecosystem - The CXL protocol is being adopted by leading CPU manufacturers, with Intel and AMD launching compatible CPUs, and companies like SK Hynix and Samsung developing CXL-compatible memory modules [11][8] - The introduction of CXL Switch technology by Alibaba Cloud marks a significant advancement in server architecture, allowing for efficient memory resource sharing [20][10] Memory Pool Construction - CXL can create memory pools that address the challenges of memory islands and data overflow in data centers, enhancing memory utilization and performance [14][12] - The CXL protocol includes three core sub-protocols that facilitate memory pooling and management, allowing for dynamic memory allocation across multiple servers [15][16] Future Demand Projections - By 2030, the demand for CXL Switch and CXL MXC chips is projected to reach 7.06 million and 64.52 million units, respectively, under optimistic scenarios [27][28] - The report suggests that as the CXL ecosystem matures, the shipment volumes of CPUs and servers compatible with CXL protocols will gradually increase [29] Investment Recommendations - The report recommends focusing on companies such as 澜起科技 (CXL MXC chip supplier), 江波龙 (CXL AIC expansion card supplier), 佰维存储 (CXL 2.0 DRAM memory expansion module supplier), and 聚辰股份 (VPD chip supplier for CXL modules) as potential investment opportunities [29][5]

Core Viewpoint - The demand for memory bandwidth and capacity is continuously increasing in the AI era, highlighting mismatches between memory supply and computational node demand, as well as low memory utilization efficiency. CXL technology is expected to build memory pools, achieving "storage-compute decoupling" and increasing penetration rates as the CXL ecosystem matures [1]. Group 1: CXL Technology and Ecosystem - Server-grade CPUs are increasingly adapting to the CXL protocol, with Intel and AMD releasing CPUs that support CXL 2.0. SK Hynix and Samsung are actively developing and mass-producing CXL-compatible memory modules. Alibaba Cloud, in collaboration with Intel, has launched the world's first server dedicated to PolarDB databases based on CXL 2.0 Switch technology [1]. - CXL technology addresses the "storage-compute imbalance" in the AI era by constructing memory pools, allowing for unified pooling of memory resources across CPUs, accelerators, and storage devices, thus enabling efficient cross-device collaboration and resource optimization [1]. - CXL technology allows for memory capacity expansion beyond local memory by over 10 times through memory expanders, meeting memory demands without increasing the number of physical servers [1]. Group 2: Future Demand and Investment Recommendations - By 2030, the demand for CXLSwitch and CXLMXC chips is expected to significantly increase, with optimistic projections estimating demand to reach 70.6 million and 64.52 million units, respectively [2]. - Investment recommendations include关注澜起科技 (CXLMXC chip supplier), 江波龙 (301308) (CXLAIC expansion card supplier), 佰维存储 (CXL 2.0 DRAM memory expansion module supplier), and 聚辰股份 (CXL module VPD chip supplier) [2].

Core Insights - Intel and Alibaba Cloud announced a series of cloud instances and storage solutions based on the new generation Xeon® 6 processors, addressing the challenges posed by AI scalability on cloud infrastructure [1][9] - High performance, high elasticity, and low total cost of ownership (TCO) are becoming key competitive indicators for global cloud providers [1] Group 1: Cloud Infrastructure Innovations - The introduction of "memory pooling" and flexible architecture is transforming cloud infrastructure, allowing dynamic allocation of resources based on demand [2][6] - Alibaba Cloud has deployed a unified hardware architecture across 29 global regions and 91 availability zones, enabling rapid resource allocation in response to sudden computing demands [4][9] Group 2: AI and Heterogeneous Computing - AI-driven heterogeneous computing is redefining the role of CPUs as central coordinators, with Intel integrating AMX matrix acceleration instruction sets to support various precision calculations [7] - The Xeon® 6 processors can efficiently handle large AI models, demonstrating significant performance improvements in various applications, such as data preprocessing for autonomous driving [7][8] Group 3: Collaboration and Competitive Edge - The stability and engineering support of the collaboration between Intel and Alibaba Cloud are highlighted as foundational elements for their long-term partnership [8] - The optimization of both hardware and software is becoming a key differentiator in the market, with Alibaba Cloud leveraging CXL 2.0 memory pooling technology for enhanced performance [8][9] Group 4: Future Directions - The shift from cloud adoption to intelligent cloud solutions is seen as an inevitable development path, with AI moving into a phase of large-scale application [9][10] - The collaboration between Intel and Alibaba Cloud aims to provide scalable and sustainable pathways for various industries through enhanced hardware performance and optimized software stacks [9][10]

Core Insights - CXL technology is enhancing computing efficiency through storage pooling and high-speed interconnects, with significant implications for AI applications [1] - Major players like NVIDIA and Alibaba Cloud are actively developing CXL capabilities to improve performance and resource utilization in AI systems [2][3] Group 1: CXL Technology Overview - CXL is an open high-speed serial protocol designed to facilitate communication between CPU, memory, and GPU, achieving higher data throughput and lower latency [1] - The technology supports efficient collaboration between accelerators like GPUs and FPGAs with main processors, addressing memory bandwidth bottlenecks and enhancing computational efficiency [1] Group 2: NVIDIA's Strategic Moves - NVIDIA has invested $5 billion in Intel to develop customized x86 CPUs for its AI infrastructure, leveraging Intel's role in the CXL alliance to enhance interoperability between NVLink and CXL technologies [2] - The acquisition of Enfabrica allows NVIDIA to integrate advanced AI interconnect technologies, including low-latency data paths and high-capacity memory support, optimizing GPU and CPU interconnects [2] Group 3: Alibaba Cloud's Innovations - Alibaba Cloud has launched the world's first CXL 2.0 Switch-based PolarDB database server, achieving ultra-low latency and high bandwidth for remote memory access [3] - The server enhances resource utilization and inference throughput by enabling collaborative pathways between GPU, CPU, and shared memory pools, positioning itself as a robust foundation for AI-driven data solutions [3]