Group 1 - The report highlights a strong outlook for the semiconductor industry, particularly in the storage chip sector, driven by multiple factors including AI demand and smartphone upgrades [3][49][62] - The DeepSeek AI model has initiated a new wave of innovation in the industry, leading to significant market interest and investment opportunities [9][14][24] - The report notes a significant divergence in performance among electronic sub-sectors in 2025, with printed circuit boards and integrated circuit manufacturing showing substantial gains, while consumer electronics lagged [3][8][25] Group 2 - The three major storage giants, Micron, Samsung, and SK Hynix, have initiated price increases for their products, indicating a new growth cycle in the storage market [51][58] - Historical analysis shows that the storage chip market has experienced cyclical trends, with the current cycle being driven by both AI server demand and smartphone upgrades [62][68] - The report emphasizes the importance of the semiconductor industry in China, particularly in overcoming technological barriers and achieving breakthroughs in critical areas [3][29][47] Group 3 - The report identifies key investment opportunities in the semiconductor supply chain, particularly in storage chip modules, packaging, and manufacturing, as well as related equipment and materials [4][49][62] - The impact of U.S.-China trade policies is highlighted, with certain sectors benefiting from export restrictions while consumer electronics face challenges [25][29][31] - The global server market is projected to grow significantly, with China playing a crucial role in driving this growth [33][34][62]

Group 1 - The stock price of Jiangbolong (301308.SZ) closed at 297.08 yuan, down 5.69% on March 31 [1] - Jiangbolong has had only one brokerage report this year, published by China Merchants Securities on March 5, titled "Jiangbolong Deep Report: Domestic Memory Leader's Multi-Dimensional Layout Welcomes Broad Growth Space Along with AI Trend" [1]

Core Viewpoint - The article emphasizes that artificial intelligence (AI) is becoming a reality, with significant investments in infrastructure and models, and anticipates that 2026 will mark a year of large-scale AI deployment, particularly in edge applications [1][2]. Group 1: AI Storage Needs - The core tasks of storage systems in AI training focus on handling massive data throughput and high-frequency checkpoint writes to prevent I/O bottlenecks, which has led to a demand for high bandwidth memory (HBM) and large SSDs [3]. - In edge AI applications, the focus shifts to inference, requiring innovations in power consumption, performance, and size due to the close integration with application scenarios [3][4]. - The need for efficiency in both training and inference is highlighted, with a call for a layered approach to storage processing to address high costs and token expenses in edge AI [3][4]. Group 2: Customized Storage Solutions - Edge AI requires deeply integrated, customized storage solutions rather than generic products, focusing on high-performance capacity and system-level integration [5][7]. - The company has developed a comprehensive capability across the entire supply chain, including chip design and manufacturing, to provide tailored storage services for edge AI applications [7][8]. Group 3: Product Innovations - The company showcased its new PCIe Gen5 mSSD, designed for edge AI devices, featuring a compact size and high performance, with read/write speeds reaching up to 11GB/s and 10GB/s, respectively [9][10]. - The mSSD's innovative cooling solution allows for sustained high performance, significantly improving thermal management compared to conventional SSDs [13][14]. Group 4: Intelligent Storage Solutions - The introduction of the Storage Processing Unit (SPU) and the Intelligence Storage Agent (iSA) creates a synergistic hardware-software ecosystem for edge AI storage, enhancing storage scheduling efficiency [16][19]. - The SPU, designed specifically for AI applications, balances capacity and cost, offering significant advantages over traditional storage solutions [16][19]. Group 5: Advanced Caching Technology - The High Level Cache (HLC) technology integrates with SPU to optimize performance and cost in edge AI devices, allowing for efficient data management and reduced DRAM requirements [21][22]. - The HLC technology has demonstrated significant performance improvements in real-world applications, achieving response times comparable to higher-capacity configurations [22]. Group 6: System in Package (SiP) Technology - The company has developed a complete SiP design process, enabling the integration of multiple chips into a single package, which is crucial for compact edge AI devices [25][26]. - This technology not only reduces hardware size but also enhances thermal management and structural layout, making it a competitive solution for various edge AI applications [26]. Conclusion - The advancements in storage technology and the strategic focus on edge AI applications position the company as a leader in the evolving landscape of AI, emphasizing the importance of innovation and collaboration in driving industry growth [28].

Group 1: Supply Chain and Partnerships - The company has established long-term strategic partnerships with multiple wafer manufacturers, including Yangtze Memory Technologies and ChangXin Memory Technologies, ensuring a stable supply chain [3] - Current wafer inventory and supply arrangements are sufficient to meet future business development and customer order demands, with plans to sign larger wafer procurement agreements in the future [3] Group 2: AI Storage Market Impact - New AI applications like OpenClaw are expected to significantly enhance the intelligence of end devices, necessitating higher capacity, bandwidth, and lower latency storage products [3] - The company has launched several new embedded products, including UFS4.1, ultra-thin ePOP5x, and ultra-small eMMC, which are entering the supply chains of major global smartphone manufacturers and are projected for large-scale commercial application in 2026 [3][4] Group 3: Product Development and Standards - The company has developed mSSD products using advanced system-in-package (SiP) technology, meeting high PCIe performance standards and low power requirements of NVMe, gaining recognition from leading PC manufacturers [4] - Efforts are underway to position mSSD as the new standard for AI edge applications, further solidifying the company's leading position in the AI edge market [4] Group 4: Investor Relations Activity - The investor relations activity involved specific targeted research with various asset management firms and securities companies, indicating a proactive approach to stakeholder engagement [2] - The event did not involve the disclosure of any undisclosed significant information [4]

Core Insights - The storage semiconductor industry is undergoing a transformation driven by AI, with NAND flash memory expected to be the largest application market by 2026 due to the increasing demand for AI inference and eSSD technology [3][15][18] - AI server storage usage is projected to be 4-5 times that of general servers, with significant growth expected in various memory types, including SOCAMM, NAND, DRAM, and HBM [3][21] - The supply side is facing structural mismatches, leading to a persistent shortage of storage capacity until 2027-2028, with a shift from aggressive expansion to disciplined production [4][30] Demand Side - AI is rapidly consuming storage semiconductor capacity, with AI server storage expected to account for over 50% of DRAM usage by 2026 [3][21] - The demand for eSSD is driven by AI inference, with projections indicating that server applications will account for 37% of NAND usage by 2026 [3][21] - The growth of KV cache is exponential, with estimates suggesting that a single inference session could generate approximately 45PB of KV cache, necessitating high-performance eSSD solutions [3][21] Supply Side - The storage industry is experiencing a structural mismatch, with limited capacity growth expected from 2026 to 2027, leading to ongoing shortages [4][30] - Inventory levels are dropping below historical safety lines, with the industry facing a significant backlog in consumer-grade production [4][30] - The overall NAND bit shipment growth is projected to remain between 10%-20% from 2023 to 2026, indicating a cautious approach to capacity expansion [4][30] Price Trends - Storage prices are expected to maintain an upward trend throughout 2026, with a gradual convergence in price increases anticipated [4][30] - The current price surge is characterized as a long-term trend rather than a cyclical rebound, with significant increases observed in contract and spot prices [4][30] New Products and Innovations - The storage technology landscape is shifting from micro-innovations to system-level restructuring, with new technologies like CXL, CIM, and PNM expected to move from concept to commercial scale in the next 2-3 years [5][30] - Major companies are launching high-performance eSSD products, with capacities reaching up to 256TB, aimed at meeting the demands of AI applications [5][30] Investment Recommendations - The storage sector is poised for significant performance releases in 2026, driven by the interplay of price and demand dynamics [6][30] - Companies to watch include major international players like SanDisk, Micron, SK Hynix, and domestic manufacturers such as Changxin Technology and Yangtze Memory Technologies [6][30]

证券代码：301308 证券简称：江波龙 公告编号：2026-026 深圳市江波龙电子股份有限公司 关于为子公司提供担保的进展公告 本公司及董事会全体成员保证信息披露的内容真实、准确、完整，没有虚假 记载、误导性陈述或重大遗漏。 特别风险提示： 深圳市江波龙电子股份有限公司（以下简称"公司"）提供的担保总额超过 公司最近一期经审计净资产的 100%，其中对资产负债率超过 70%的单位的担保 总额超过公司最近一期经审计净资产 50%，前述担保全部为公司对合并报表范围 内子公司的担保，敬请投资者充分关注担保风险。 一、担保情况概述 公司于 2025 年 3 月 19 日、2025 年 4 月 11 日分别召开了第三届董事会第八 次会议、第三届监事会第八次会议以及 2024 年年度股东大会，会议审议通过了 《关于公司 2025 年度担保额度预计的议案》。为了满足公司 2025 年度日常经营 及业务发展所需，自 2024 年年度股东大会审议通过之日起至 2025 年年度股东会 召开之日止，公司拟为子公司向银行等金融机构申请综合授信或其他日常经营所 需，提供总额不超过人民币 110 亿元（或等值外币）的担保额度。 注： ...

Core Viewpoint - The company has not established direct cooperation with Nvidia but is actively developing AI-related storage products for various applications [1] Group 1: Product Development - The company has launched enterprise-level storage products including SATA SSD, PCIe SSD, and RDIMM, focusing on AI large model application scenarios [1] - New high-performance storage products such as MRDIMM and CXL2.0 memory expansion modules have been introduced [1] - The product offerings cater to a diverse clientele, including telecom operators, internet companies, and server manufacturers [1]

Core Viewpoint - The article discusses the recent decline in the A-share storage chip market, highlighting the impact of Google's new compression algorithm, TurboQuant, on memory demand and the overall storage industry outlook [1][3][5]. Group 1: Market Performance - On March 27, A-share storage chips experienced a collective decline, with companies like Baiwei Storage dropping over 6% at one point, and other firms such as Jiangbolong and Zhaoyi Innovation also seeing declines of over 3% [1][2]. - The storage index showed significant drops across various companies, with Jiangbolong at -3.73%, Zhongdian Port at -3.31%, and Baiwei Storage at -3.01% [2]. Group 2: Technological Developments - Google introduced TurboQuant, a compression algorithm aimed at reducing memory requirements for AI systems, particularly addressing key-value cache bottlenecks [3][4]. - TurboQuant can compress key-value caches to 3-bit precision without retraining models, achieving approximately 6 times memory compression while maintaining model accuracy [3][4]. Group 3: Industry Outlook - Despite the recent market downturn, the long-term demand for storage chips remains strong, driven by increased investments in AI infrastructure by overseas CSPs, which is expected to boost enterprise storage demand [5]. - Analysts predict that the current storage cycle's intensity and sustainability may surpass the previous cycle, with significant profit increases for related industry chain companies [5]. - Supply constraints are expected to persist until at least the second half of 2027, despite major DRAM manufacturers planning production increases of about 26% for DRAM and 24% for NAND by 2026 [5].

Core Viewpoint - The article discusses the recent performance of semiconductor stocks in the A-share market, highlighting the impact of AI demand on the semiconductor industry and the differentiation in the storage chip sector [3][5][6]. Group 1: Semiconductor Stock Performance - Multiple semiconductor stocks in the A-share market saw gains, with Jin Hai Tong (603061.SH) rising over 4%, and other companies like SiDa Semiconductor (603290.SH) and HaiGuang Information (688041.SH) also experiencing increases [3]. - The Philadelphia Semiconductor Index in the U.S. rose by 1.21%, with Arm's stock surging by 16.38% after announcing its first self-developed AGI CPU, projected to generate $15 billion in revenue per product by 2031 [4]. Group 2: AI Demand and Market Dynamics - Intel and AMD have informed customers of upcoming price increases for CPUs in March and April, indicating a growing demand for AI computing power and an expanding market space [5]. - Google introduced a new AI memory compression technology, TurboQuant, which can reduce memory usage for large language models by at least six times without sacrificing accuracy, raising concerns about future storage demand [5]. Group 3: Storage Chip Sector Analysis - The growth in the storage chip sector is entering a phase of differentiation, with AI demand expected to provide long-term market growth momentum [6][7]. - TrendForce data indicates that the global supply-demand imbalance for memory will continue into 2026, with major DRAM manufacturers expected to increase production by approximately 26% and NAND production by about 24% [7]. - Despite the anticipated production increases, the actual expansion of supply from domestic manufacturers may take time, leading to a potential shift from broad sector growth to structural differentiation, particularly favoring high-end markets like HBM [7].

证券代码：301308 证券简称：江波龙 公告编号：2026-025 深圳市江波龙电子股份有限公司 一、担保情况概述 公司于 2025 年 3 月 19 日、2025 年 4 月 11 日分别召开了第三届董事会第八 次会议、第三届监事会第八次会议以及 2024 年年度股东大会，会议审议通过了 《关于公司 2025 年度担保额度预计的议案》。为了满足公司 2025 年度日常经营 及业务发展所需，自 2024 年年度股东大会审议通过之日起至 2025 年年度股东会 召开之日止，公司拟为子公司向银行等金融机构申请综合授信或其他日常经营所 需，提供总额不超过人民币 110 亿元（或等值外币）的担保额度。 公司于 2026 年 3 月 2 日、2026 年 3 月 19 日分别召开了第三届董事会第十 六次会议、2026 年第一次临时股东会，会议审议通过了《关于公司 2025 年度新 增对子公司担保额度的议案》，在上述 2024 年年度股东大会审议通过的担保额 度基础上，公司新增对子公司提供不超过 40.55 亿元人民币或其他等值币种的担 保额度。综上，公司 2025 年度对子公司担保总额度为人民币 150.55 亿元。 ...