In-Depth Research on the Storage Chip Industry Chain 1. Introduction: Overview of the Storage Chip Industry 1.1 Strategic Position and Classification System of Storage Chips Storage chips, as a core component of the semiconductor industry, hold an irreplaceable strategic position in the digital economy era. According to IC Insights, storage chips account for approximately 56% of the entire semiconductor market, with DRAM accounting for about 56%, NAND Flash accounting for about 43%, and other storage chips (EEPROM, EPROM, ROM, SRAM) collectively accounting for about 1%. Storage chips are mainly divided into three categories: DRAM (Dynamic Random Access Memory), NAND Flash (Flash Memory), and NOR Flash (Read-Only Memory). DRAM, as the main memory, interacts directly with the CPU via the bus, storing data and intermediate computation results, and is currently developing towards advanced 3D stacking architectures. NAND Flash is a non-volatile memory that stores data in the form of electric charges, primarily used in solid-state drives, embedded storage, and mobile storage. NOR Flash is suitable as a storage medium for executing code, used in scenarios requiring fast code reading. 1.2 Industry Chain Structure and Technological Evolution Trends The storage chip industry chain is divided into four major links: upstream raw materials and equipment, midstream chip design and manufacturing, downstream packaging testing and module integration, and terminal application scenarios. The upstream includes semiconductor material suppliers such as silicon wafers, photoresists, target materials, polishing materials, and electronic specialty gases, as well as semiconductor equipment suppliers for lithography, etching, thin film deposition, cleaning, and packaging testing. In terms of technological evolution, storage chips are developing towards higher density, larger capacity, and lower power consumption. DRAM technology is evolving from DDR4 to DDR5 and LPDDR5X, with DDR5 starting speeds of 4800MT/s and reaching up to 8400MT/s, a significant improvement compared to DDR4’s 3200MT/s. NAND technology is evolving from planar NAND to 3D NAND, with the number of stacked layers continuously increasing; companies like Samsung and Kioxia have achieved technological breakthroughs of over 300 layers. 1.3 Global Industry Landscape and Competitive Situation The global storage chip industry presents a highly concentrated oligopolistic structure. In the DRAM field, Samsung, SK Hynix, and Micron together account for over 90% of the global market share. In the NAND Flash field, a stable market structure is formed by five major manufacturers: Samsung, Kioxia, Western Digital, Micron, and SK Hynix. In recent years, China’s storage chip industry has developed rapidly. Yangtze Memory Technologies has become the only domestic company capable of producing 3D NAND Flash, with a production capacity expected to reach 150,000 wafers per month by 2025, accounting for about 8% of global NAND Flash supply. Changxin Memory focuses on DRAM manufacturing, with a production capacity expected to reach 120,000 wafers per month by 2025, with technology catching up to the international mainstream of 19nm. The domestic self-sufficiency rate of storage chips has increased from 5% in 2020 to 18% in 2025. 2. Technical Specifications and Application Analysis of Storage Chips 2.1 DRAM Technical Specifications and Application Scenarios 2.1.1 Generational Evolution and Performance Parameters of DRAM Technology DRAM technology has undergone multiple generations of evolution from DDR1 to DDR5, with each generation achieving significant breakthroughs in speed, capacity, and power consumption. DDR5, as the fifth generation of double data rate synchronous dynamic random access memory, has a starting speed of 4800MT/s, while DDR4’s maximum speed is 3200MT/s, with DDR5 planning to increase the standard speed to 8800MT/s or even higher. DDR5 has significant architectural improvements over DDR4. DDR5 divides the memory module into two independent 32-bit addressable sub-channels to improve the efficiency of data access by the memory controller and reduce latency. In terms of capacity, DDR5 can reach a maximum single-module capacity of 128GB, four times that of DDR4, and can be expanded to 512GB in a four-channel configuration. In terms of power consumption, DDR5 operates at a voltage of 1.1V, lower than DDR4’s 1.2V. The low-power DRAM (LPDDR) series is specifically designed for mobile devices. LPDDR5 operates at a voltage further reduced to 1.05V, with data transfer rates reaching 6400-8533MT/s. The latest LPDDR5X technology can achieve data rates of up to 10667Mbps, with performance improved by 25% and power consumption reduced by 25% compared to the previous generation. High Bandwidth Memory (HBM) represents the highest level of DRAM technology. HBM achieves bandwidth and capacity far exceeding traditional DDR or GDDR by vertically stacking multiple DRAM chips and combining through-silicon via (TSV) technology. HBM3 achieves a bandwidth of 819GB/s, HBM3E further increases it to 1.2TB/s, while HBM4 aims to achieve a bandwidth of 2TB/s. 2.1.2 Application Scenarios and Performance Requirements of Different Specifications of DRAM Different specifications of DRAM play a critical role in various electronic devices, with significant differences in performance requirements and application scenarios. In the smartphone sector, DRAM is used as the operating memory to ensure multitasking. With the popularization of 5G and AI features, the demand for DRAM in smartphones is rapidly increasing. High-end flagship smartphones typically come with 12-16GB of LPDDR5 or LPDDR5X memory, mid-range smartphones are equipped with 8-12GB of LPDDR4X or LPDDR5 memory, while entry-level smartphones are equipped with 4-8GB of LPDDR4X memory. In the personal computer sector, DRAM is the core of computer memory modules. Desktop and laptop computers typically use DDR4 or DDR5 memory, with capacities ranging from 8GB to 64GB. Gaming computers and workstations require larger capacities and higher frequency memory, usually equipped with 16-64GB of DDR4-3200 or DDR5-4800 and above. Servers and data center devices have the most demanding requirements for DRAM. Traditional servers typically come with 128-512GB of DDR4 or DDR5 memory, while AI servers have exponentially increasing demands. A single AI server requires 2-4TB of DDR5 memory and 640GB of HBM, which is eight times that of ordinary servers. In the gaming console sector, DRAM supports smooth game operation. The Sony PlayStation 5 is equipped with 16GB of GDDR6 memory, the Microsoft Xbox Series X has 16GB of GDDR6 memory, while the Nintendo Switch uses 4GB of LPDDR4 memory. 2.2 NAND Technical Specifications and Application Scenarios 2.2.1 NAND Storage Technology Routes and Performance Characteristics NAND Flash is classified into four types based on the number of bits stored in each storage cell (Cell): SLC (Single-Level Cell), MLC (Multi-Level Cell), TLC (Triple-Level Cell), and QLC (Quad-Level Cell). SLC stores 1bit of data per Cell, MLC stores 2bits, TLC stores 3bits, and QLC stores 4bits. In terms of performance, SLC has the highest read and write speeds and the longest lifespan, with a theoretical erase cycle of up to 100,000 times, but at the highest cost. MLC achieves a good balance between performance and cost, with a theoretical erase cycle of 3000-5000 times. TLC further increases storage density and lowers costs, with a theoretical erase cycle of 1000-3000 times, becoming the market mainstream. QLC has the highest storage density and the lowest cost, but the shortest lifespan, with a theoretical erase cycle of only 150-300 times. The development of 3D NAND technology has greatly improved storage density and performance. Samsung’s 8th generation V-NAND has reached 236 layers, using TLC Flash; Kioxia and Western Digital have launched 218-layer 3D NAND Flash, planning to introduce technology with over 500 layers by 2032. 3D NAND significantly increases capacity by vertically stacking storage cells without increasing chip area. 2.2.2 Application Scenarios and Capacity Configurations of Different Types of NAND NAND Flash is widely used in the storage systems of various electronic devices, with differences in capacity configurations and application scenarios among different types of NAND. In the smartphone sector, NAND is used as onboard storage for operating systems, applications, and user data. Smartphones typically use NAND chips with UFS (Universal Flash Storage) interfaces. Entry-level smartphones are equipped with 64-128GB of storage, mid-range smartphones with 128-256GB, and high-end flagship smartphones with 256-1TB of storage. Some Chinese smartphone manufacturers plan to adopt QLC UFS solutions starting in the fourth quarter of 2024, while Apple is expected to introduce QLC Flash in iPhones by 2026. In the personal computer sector, NAND is primarily used in solid-state drives (SSDs). Consumer-grade SSDs typically use TLC or QLC NAND, with capacities ranging from 128GB to 8TB. High-end SSDs are equipped with independent DRAM caches to enhance performance, while entry-level SSDs rely on host memory or no-cache designs. Enterprise-grade SSDs typically use MLC or SLC NAND to ensure reliability, with capacities reaching 16TB or more. In the server and data center sector, NAND is used to store massive amounts of data. SSDs used in data centers typically employ enterprise-grade MLC or TLC NAND, equipped with large-capacity DRAM caches and advanced error correction algorithms. With the growth of AI and big data applications, the demand for high-capacity, high-performance SSDs in data centers is rapidly increasing, with single server SSD configurations upgrading from 64TB to 96TB. In the gaming console sector, NAND is used to store game installation packages, save files, and system data. The Sony PS5 is equipped with a custom 825GB SSD, the Microsoft Xbox Series X has a custom 1TB SSD, both using high-speed NVMe interfaces and high-performance NAND chips. 2.3 Overview of Other Types of Storage Chips In addition to DRAM and NAND, there are other types of storage chips that play important roles in specific fields. NOR Flash is a type of read-only memory suitable as a storage medium for executing code, used in scenarios requiring fast code reading. The main advantage of NOR Flash is that it can execute code directly (XIP, eXecute In Place) without first loading the code into RAM, making it widely used in embedded systems. The global NOR Flash market is dominated by manufacturers such as Winbond, Macronix, and GigaDevice, with GigaDevice’s Serial NOR Flash market share rising to the second position globally. EEPROM (Electrically Erasable Programmable Read-Only Memory) is a type of storage chip that retains data after power loss and can be erased and rewritten multiple times. EEPROM is mainly used to store configuration information, calibration data, and other small amounts of critical information, with erase cycles typically exceeding 100,000 times. New storage technologies are also continuously developing. Resistive Random Access Memory (RRAM) is considered a next-generation storage technology with advantages of high speed, low power consumption, and high density, listed as a key technology for breakthroughs. Phase Change Memory (PCM), Magnetic RAM (MRAM), and other technologies are also seeking breakthroughs in specific fields. 3. Analysis of Production Capacity Distribution of Major Application Devices 3.1 Smartphone Production Capacity Distribution Smartphones are one of the largest application markets for DRAM and NAND storage chips. According to TrendForce, global smartphone production is expected to reach approximately 1.21 billion units in 2024, requiring a large number of DRAM and NAND storage chips. In terms of production capacity distribution, China is the core of global smartphone manufacturing, accounting for 67% of global capacity. China is not only the production base for domestic brands such as Huawei, Xiaomi, OPPO, and vivo but also the core of contract manufacturing for international brands like Apple. Major contract manufacturers such as Foxconn and Pegatron have extensive production networks in China, handling most of the global smartphone assembly tasks. India is the second-largest base for global smartphone manufacturing, accounting for 12% of global capacity. In recent years, with the promotion of the “Make in India” policy and labor cost advantages, more and more smartphone manufacturers are establishing production bases in India. Samsung’s factory in India is continuously expanding, and Chinese brands like Xiaomi, OPPO, and vivo are also investing heavily in factories in India. Vietnam is another important smartphone manufacturing base, accounting for 10% of global capacity. Vietnam attracts capacity transfers from Korean companies like Samsung and LG, as well as some Chinese brands, due to its geographical advantages, relatively low labor costs, and improving infrastructure. South Korea mainly produces high-end Samsung models, contributing 5% of global capacity. Samsung’s factories in Hwaseong and Giheung primarily produce high-end models such as the Galaxy S and Galaxy Note series, employing the most advanced production technologies and quality management systems. 3.2 Personal Computer Production Capacity Distribution Personal computers, including desktops, laptops, and tablets, are an important application area for storage chips. According to IDC, global PC shipments are expected to reach approximately 280 million units in 2024, with laptops accounting for over 70%. China is the absolute core of global PC manufacturing, accounting for over 75% of global capacity. Lenovo, as the world’s largest PC manufacturer, keeps 90% of its production capacity in China. International brands like Dell and HP still rely on Chinese factories for about 75% of their production capacity, mainly concentrated in the Yangtze River Delta and Pearl River Delta regions. Major contract manufacturers such as Foxconn, Quanta, Compal, and Wistron have large-scale production bases in China. Southeast Asia has taken on about 20% of PC production capacity, mainly concentrated in countries like Thailand and Vietnam. These regions are primarily responsible for the assembly of mid-range and low-end PCs, with Apple’s MacBook and some HP product lines produced in these areas. The United States retains some high-end PC and workstation manufacturing capacity, mainly produced by domestic factories of companies like Dell and HP. These factories typically employ highly automated production lines to produce customized and high-end products. 3.3 Solid State Drive Production Capacity Distribution The core component of solid-state drives (SSDs) is NAND Flash chips, with some high-end products also paired with DRAM chips to enhance performance. According to TrendForce, global SSD shipments are expected to exceed 400 million units in 2024, with a market size exceeding $60 billion. The global SSD production capacity presents a triangular pattern centered on China, South Korea, and Japan, with these three countries accounting for 76% of global NAND Flash production capacity. China has rapidly risen in the SSD manufacturing field. Yangtze Memory Technologies is the only domestic company capable of producing 3D NAND Flash, with a production capacity expected to reach 150,000 wafers per month by 2025. Brands like Zhaoxin (a subsidiary of Yangtze Memory Technologies) occupy an important position in the domestic market. At the same time, international brands like Samsung, Western Digital, and Kioxia also have packaging and testing factories in China. South Korea is an important base for global SSD manufacturing, with Samsung and SK Hynix having complete NAND Flash production capabilities in South Korea. Samsung’s factories in Hwaseong and Pyeongtaek produce high-end SSD products, while SK Hynix focuses on the enterprise SSD market. Kioxia (formerly Toshiba Memory) is the second-largest NAND Flash manufacturer globally, with large-scale production bases in Yokkaichi and Iwate, Japan. Kioxia’s joint venture with Western Digital occupies an important position in the global SSD market. Malaysia and Vietnam have taken on some SSD module packaging and testing capacity, but their participation in core chip manufacturing is relatively low. 3.4 Server and Data Center Equipment Production Capacity Distribution Servers and data center equipment are the fastest-growing areas for storage chip demand, especially with the explosive growth of AI servers requiring large-capacity DRAM and high-performance NAND. China is an important base for global server manufacturing, with local manufacturers like Inspur, Huawei, and Sugon. Inspur is the third-largest server manufacturer globally, and Huawei is rapidly rising in the high-end server market. These companies’ production bases in China not only meet domestic demand but also export significantly overseas. The United States has high-end server manufacturing capacity from international server giants like Dell and HP. These companies maintain R&D and high-end manufacturing capabilities in the U.S. while having assembly factories in various locations worldwide. The U.S. leads in data center equipment manufacturing technology, especially in high-end servers, networking equipment, and storage systems. South Korea, leveraging the storage chip advantages of companies like Samsung and SK Hynix, produces related servers and data center equipment. Samsung not only produces storage chips but also provides complete server and storage solutions. Southeast Asian countries like Singapore and Malaysia have taken on some server assembly and testing capacity, mainly serving the data center construction needs in the Asia-Pacific region. 3.5 Game Console Production Capacity Distribution The demand for storage chips in game consoles mainly involves DRAM and NAND, with DRAM supporting smooth game operation and NAND used for storing game installation packages and save data. China is the core of global game console manufacturing, accounting for over 80% of global capacity. Most mainstream game consoles from Sony, Microsoft, and Nintendo are assembled by Chinese contract factories, mainly including Foxconn, Pegatron, and Flex. These contract factories are primarily concentrated in Guangdong and Jiangsu provinces. Japan retains a small amount of core component production and assembly capacity for high-end game consoles, mainly producing limited edition and high-end models. Sony’s factories in Japan are responsible for the R&D and production of some high-end products in the PlayStation series. Southeast Asia has taken on about 15% of game console assembly capacity, mainly concentrated in countries like Vietnam and Thailand, primarily responsible for the production of mid-range game consoles and accessories. The U.S. mainly handles the design and R&D of game consoles and the manufacturing of some high-end components, such as Microsoft’s Xbox series, which has R&D centers and some production facilities in Washington State. 4. Global Distribution Pattern of Storage Chip Production Capacity 4.1 Production Capacity Layout of Major Manufacturers Global storage chip production capacity is highly concentrated in a few large enterprises, which optimize costs, mitigate risks, and stay close to markets through globalization. In the DRAM field, Samsung, SK Hynix, and Micron have formed a highly monopolistic structure. Samsung has production bases in Hwaseong, Pyeongtaek, and Xi’an, China, with the Xi’an factory’s monthly capacity reaching 250,000 wafers, accounting for 40% of its total NAND production capacity. SK Hynix has production facilities in Icheon, Wuxi, China, and Oregon, USA, with the Wuxi factory being an important NAND production base. Micron’s production bases are distributed in Utah, Idaho, Japan’s Hiroshima, and Singapore, with the Fab 7 factory in Singapore being its largest NAND production base, with a monthly capacity slightly above 300,000 wafers. In the NAND Flash field, in addition to Samsung, SK Hynix, and Micron, Kioxia (formerly Toshiba Memory) and Western Digital are also important players. Kioxia has factories in Yokkaichi, Iwate, Japan, and Dalian, China, with the Dalian factory being an important overseas production base. Western Digital has joint ventures with Kioxia and has assembly and testing factories in Malaysia and Thailand. China’s storage chip industry is rapidly rising. Yangtze Memory Technologies has three 12-inch wafer fabs in Wuhan, with a total production capacity expected to reach 150,000 wafers per month by 2025, with a domestic equipment localization rate of 45%. The first fully localized production line is expected to enter trial production in the second half of 2025. Changxin Memory has a production base in Hefei, with DRAM production capacity expected to reach 120,000 wafers per month by 2025, with technology catching up to the international mainstream of 19nm. 4.2 Capacity Shares and Technological Levels of Various Countries/Regions From the perspective of global capacity distribution, Asian countries dominate, especially East Asia, which concentrates most of the global storage chip production capacity. China is the largest storage chip producer globally, not only dominating in terminal product manufacturing but also rapidly catching up in chip manufacturing. According to TrendForce, China accounts for about 15% of global NAND Flash production capacity and about 5% of DRAM production capacity. However, it is important to note that China’s self-sufficiency rate for storage chips remains low, expected to be around 18% by 2025, with most high-end storage chips still relying on imports. South Korea is one of the countries with the most advanced storage chip technology globally, holding an absolute advantage in the DRAM field. Samsung and SK Hynix together account for over 70% of global DRAM production capacity and also hold an important position in NAND. South Korean companies lead in technological innovation, with Samsung’s 1α/1β processes and SK Hynix’s HBM technology being at the forefront of the industry. Japan holds an important position in the NAND Flash field, with Kioxia being the second-largest NAND Flash manufacturer globally. Japanese companies have advantages in materials and equipment, with firms like Shin-Etsu Chemical and JSR occupying important positions in key materials such as photoresists and electronic specialty gases. The United States maintains a leading position in storage chip design and high-end manufacturing, with Micron being the third-largest storage chip manufacturer globally. U.S. companies invest heavily in technology R&D, particularly in new storage technologies such as MRAM and ReRAM. 4.3 Analysis of Factors Influencing Capacity Distribution The distribution of storage chip production capacity is influenced by various factors, including technological development levels, cost structures, policy environments, and market demand. Technological development levels are the core factor determining capacity distribution. Storage chip manufacturing involves complex process technologies that require long-term technological accumulation and substantial investment. South Korean companies have established technological and scale advantages through long-term accumulation in DRAM technology. Japanese companies have traditional advantages in NAND Flash and related materials technology. U.S. companies maintain a leading position in design and R&D. Chinese companies are rapidly catching up through technology introduction and independent R&D. Cost structure is an important factor influencing capacity layout. Storage chip manufacturing requires significant capital investment and operational costs, including equipment investment, raw material procurement, energy consumption, and labor costs. There are significant differences in these cost factors across different regions. China has comprehensive advantages in labor costs, infrastructure, and industrial support, making it the most important manufacturing base globally. Southeast Asian countries have advantages in labor costs, taking on some assembly and testing capacity. Developed countries maintain advantages in high-end manufacturing and R&D but have higher costs. The policy environment has a significant impact on capacity distribution. Governments in various countries support the development of their storage chip industries through industrial policies, tax incentives, and R&D subsidies. The Chinese government strongly supports the storage chip industry through the National Big Fund and industrial policies, with the third phase of the National Big Fund having a registered capital of 344 billion yuan, focusing on storage chips and other fields. The U.S. influences global capacity distribution through export controls and investment restrictions, attempting to curb the development of China’s storage chip industry. Countries like South Korea and Japan also support their companies in maintaining technological leadership through policies. Market demand is a significant driving force for capacity distribution. The main application markets for storage chips include China, the U.S., and Europe, and capacity layouts need to consider proximity to markets to reduce logistics costs and shorten delivery cycles. Additionally, the market demand structure in different regions also affects capacity layouts, such as the high demand for mid-to-low-end storage chips in the Chinese market and the higher demand for high-end products in Europe and the U.S. 5. Analysis of Key Upstream Links in the Industry Chain 5.1 Distribution of Manufacturing Equipment Suppliers Storage chip manufacturing requires a large number of high-end equipment, which are high in technical content and value, with major suppliers concentrated in a few developed countries. Lithography machines are the most technically demanding and complex equipment in chip manufacturing, accounting for over 35% of chip manufacturing costs. The global lithography machine market is dominated by ASML, Nikon, and Canon, with ASML holding an absolute leading position, especially in the EUV lithography machine field, forming a monopoly. China’s Shanghai Micro Electronics has achieved breakthroughs in DUV lithography machines, with a production yield of 82% for 28nm immersion lithography machines, producing over 100 units annually. Etching machines are core process equipment in semiconductor patterning. The global etching machine market is dominated by LAM Research (Lam), Applied Materials, and Tokyo Electron, with LAM Research holding a 53% market share in dry etching machines. Chinese companies like Northern Huachuang and Zhongwei have made significant progress in etching equipment, with Zhongwei’s 5nm etching machine validated by TSMC and Northern Huachuang’s 14nm process equipment achieving mass production. Thin film deposition equipment, including Chemical Vapor Deposition (CVD), Physical Vapor Deposition (PVD), and Atomic Layer Deposition (ALD), is one of the key equipment in chip manufacturing. Major global suppliers include Applied Materials, Lam, and Tokyo Electron. Chinese companies like Northern Huachuang, Zhongwei, Tuojing Technology, and Weidao Nano are actively laying out in this field. Other key equipment includes cleaning equipment, ion implanters, Chemical Mechanical Polishing (CMP) equipment, and testing equipment. The suppliers of these equipment are mainly concentrated in developed countries such as the U.S., Japan, and the Netherlands. Chinese companies have achieved breakthroughs in some equipment fields, such as Shengmei Shanghai in cleaning equipment and Huahai Qingke in CMP equipment. 5.2 Key Materials Supply Chain Storage chip manufacturing requires various key materials, and the stability of their supply directly affects production capacity and costs. Silicon wafers are the basic materials for storage chips, accounting for 30%-40% of total chip costs. The global silicon wafer market presents a highly concentrated oligopolistic structure, with the top six manufacturers—Shin-Etsu Semiconductor, SUMCO, GlobalWafers, Siltronic, SK Hynix Siltronic, and Soitec—accounting for about 80% of the market share. Chinese companies like Shanghai Silicon Industry, Zhonghuan Semiconductor, and Lianwei Microelectronics have made progress in the silicon wafer field, with a localization rate of about 20-30%. Electronic specialty gases are the “blood” of chip manufacturing, used in etching, deposition, and other processes, requiring extremely high purity, usually above 6N level. Major global suppliers include U.S. Air Products, Germany’s Linde, and France’s Air Liquide. Chinese companies like Huate Gas, Jinhong Gas, and Yake Technology have achieved breakthroughs in different types of electronic specialty gases, with Huate Gas being the only domestic gas supplier certified by ASML. Photoresists are core materials in the lithography process, with very high technical content. The global photoresist market is mainly dominated by Japanese companies, including JSR, Tokyo Ohka, and Shin-Etsu Chemical. Chinese companies mainly focus on mid-to-low-end products, still relying on imports for high-end photoresists in KrF, ArF, and EUV. Target materials are used in Physical Vapor Deposition (PVD) processes to form metal interconnect layers. Major global suppliers include Honeywell and Nippon Mining & Metals. China’s Jiangfeng Electronics is a leading global supplier of ultra-high-purity aluminum/titanium/tantalum target materials, covering Yangtze Memory Technologies, SK Hynix, etc., with a market share of about 10-15% in the storage field. Other key materials include photomasks, CMP polishing materials, and electronic chemicals. The supply of these materials is highly concentrated in a few developed countries, and Chinese companies are gradually achieving domestic substitution through technological innovation and industrial upgrading. 5.3 Distribution of Design and Testing Links The storage chip industry chain also includes key links such as design and testing, which significantly impact the overall industry landscape. In the design link, global storage chip design is mainly completed by IDM (Integrated Device Manufacturer) companies, such as Samsung, SK Hynix, Micron, and Kioxia, all of which have strong design capabilities. Additionally, there are some specialized storage chip design companies, such as GigaDevice, Beijing Junzheng, and Dongxin Co., which have advantages in specific fields. In terms of design tools, EDA software is mainly monopolized by three giants: Synopsys, Cadence, and Mentor Graphics, while Chinese companies like Huada Jiutian are striving to achieve breakthroughs. In the testing link, major global testing companies include ASE, Amkor, and Powertech Technology. Chinese companies have strong capabilities in the testing field, with JCET being the third-largest testing company globally, and Tongfu Microelectronics and Huatian Technology ranking fifth and sixth globally, respectively. These companies not only provide services for domestic storage chip companies but also undertake orders from international clients. Advanced packaging technology is becoming increasingly important in the storage chip field, especially in high-end products like HBM and 3D NAND. Foundries like TSMC and Samsung are leading in advanced packaging technology, while Chinese companies like JCET and Tongfu Microelectronics are also actively laying out related technologies. The testing link is crucial for ensuring product quality, requiring specialized testing equipment and technology. Major global testing equipment suppliers include Teradyne and Advantest. Chinese companies like Changchuan Technology and Huafeng Measurement and Control have made progress in the storage chip testing equipment field. 6. Market Dynamics and Development Trends 6.1 Frontiers of Technological Development Storage chip technology is undergoing rapid evolution, with new technologies continuously emerging, driving the industry towards higher performance, lower power consumption, and larger capacity. In DRAM technology, HBM has become the most focused technology direction. HBM achieves ultra-high bandwidth through 3D stacking technology and has become standard for AI servers. HBM3E’s bandwidth reaches 1.2TB/s, HBM4 aims for 2TB/s, and HBM8 targets 64TB/s. Companies like Samsung, SK Hynix, and Micron are actively promoting the R&D and mass production of HBM technology. At the same time, DDR5 technology is also maturing, with data rates increasing from 4800MT/s to 8400MT/s, and further increases are expected. In NAND technology, 3D NAND has become the mainstream technology route, with the number of stacked layers continuously increasing. Companies like Samsung and Kioxia have achieved technological breakthroughs of over 300 layers, with future goals of 500 layers or even 1000 layers. Meanwhile, QLC technology is rapidly gaining popularity, expected to contribute to over 25% of NAND Flash bit shipments by 2025. New storage technologies like CXL (Compute Express Link) are also driving changes in storage architecture, achieving memory pooling and sharing through high-speed interconnect technology, improving resource utilization efficiency. In manufacturing processes, storage chips are advancing towards more advanced process nodes. DRAM is evolving from 1z to 1α and 1β processes, while NAND is evolving from planar to 3D and from TLC to QLC. At the same time, new materials and device structures are being continuously explored, such as GAA (Gate-All-Around) technology and new storage media. In system architecture, storage chips are deeply integrating with computing and networking technologies. New concepts like near-memory computing, storage-class memory, and intelligent storage are continuously emerging, driving changes in the entire IT architecture. Especially in the AI era, storage chips are no longer simple storage devices but have become an important part of computing systems. 6.2 Supply and Demand Relationships and Price Trends The storage chip market exhibits significant cyclical characteristics, with considerable price fluctuations. Since 2024, the storage chip market has experienced a transition from a low point to recovery. On the demand side, applications such as AI servers, smartphones, and personal computers continue to drive demand for storage chips. The explosive growth of the AI server market has led to exponential increases in demand for HBM and large-capacity DRAM. According to TrendForce, the HBM market size is expected to reach $35 billion by 2025 and $98 billion by 2030, with a compound annual growth rate of 33%. On the supply side, global storage chip production capacity is gradually recovering after adjustments in 2023, starting to recover in 2024. However, due to long equipment investment cycles and high technical barriers, supply growth is relatively slow. Additionally, geopolitical factors also affect supply stability, such as the U.S. export control policies impacting the capacity expansion of Chinese storage chip companies. In terms of price trends, storage chip prices began to rise in 2024. According to market data, the average price of DRAM increased by 53% in 2024, and NAND prices also saw significant increases. However, price trends are differentiated, with high-end products like HBM experiencing larger price increases, while ordinary DRAM and NAND prices have relatively moderate increases. It is expected that storage chip prices will continue to rise in 2025, but the growth rate may slow down. It is important to note that the supply and demand relationships and price trends in the storage chip market are influenced by various factors, including the macroeconomic environment, technological development, policy changes, and market competition, leading to significant uncertainties. 6.3 Emerging Application Drivers The growth of the storage chip market is primarily driven by emerging applications, especially the development of technologies such as AI, 5G, the Internet of Things, and autonomous driving, which have brought enormous storage demands. AI and big data applications are the most important driving forces behind the growth in storage chip demand. AI training and inference require processing massive amounts of data, placing extremely high demands on storage systems’ bandwidth, capacity, and performance. A single AI server requires 2-4TB of DDR5 memory and 640GB of HBM, which is eight times that of ordinary servers. As the scale of AI models continues to expand, the demand for storage chips will continue to grow. 5G and IoT applications have brought a large amount of edge storage demand. 5G networks require a large number of storage devices to support technologies like Network Function Virtualization (NFV) and Software-Defined Networking (SDN). The data generated by IoT devices is growing exponentially, requiring massive storage capacity. It is expected that by 2025, the number of global IoT devices will exceed 20 billion, generating over 100ZB of data. Autonomous driving and smart vehicles are another important growth area. Autonomous vehicles need to process large amounts of sensor data in real-time, placing extremely high demands on storage systems’ performance, reliability, and security. An autonomous vehicle can generate terabytes of data daily, requiring high-performance storage solutions. The development of data centers and cloud computing is also driving the growth of storage chip demand. As enterprises deepen their digital transformation, more and more data needs to be stored and processed. Cloud service providers need to build large-scale data centers equipped with a large number of servers and storage devices. It is expected that by 2025, the storage capacity of global data centers will reach over 1000EB. 7. Geopolitical and Industrial Policy Impacts 7.1 U.S. Technology Sanctions Against China The U.S. has implemented comprehensive sanctions against China’s storage chip industry, attempting to curb China’s development in this field. Since October 2022, the U.S. has placed Yangtze Memory Technologies on the Entity List, restricting U.S. companies from providing equipment and technical support to it. On December 2, 2024, the U.S. further escalated sanctions by adding high bandwidth memory (HBM) to the control list and introducing export controls on 24 types of storage chip manufacturing equipment, requiring approval for any related equipment containing U.S. technology to be exported to China. Since 2025, U.S. sanctions have further intensified, with two additions of Chinese companies to the Entity List in March and September, and a revision of terms on September 16 to include subsidiaries of Chinese companies in the sanctions scope. On November 10, the U.S. enacted new sanction rules, bringing all storage-specific equipment such as detection, etching, and cleaning under control. The U.S. has also influenced the development of China’s storage chip industry through investment restrictions. On October 28, 2024, the U.S. issued new regulations restricting U.S. companies and individuals from engaging in specific transactions with China in the semiconductor, quantum information technology, and artificial intelligence fields. In May 2025, the U.S. required three EDA giants—Synopsys, Cadence, and Siemens—to stop exporting software to China unless applying for licenses on a case-by-case basis. These sanctions have severely impacted Chinese storage chip companies, particularly in equipment procurement, technology introduction, and talent exchange. However, they have also stimulated the determination of Chinese companies to innovate independently and accelerated the process of domestic substitution. 7.2 Support for the Industry from Chinese Policies In the face of external pressures, the Chinese government has introduced a series of policy measures to support the development of the storage chip industry. In terms of financial support, the third phase of the National Big Fund was established in 2024, with a registered capital of 344 billion yuan, focusing on key areas including storage chips. The third phase of the National Big Fund is expected to leverage about 1.5 trillion yuan of social capital to provide strong financial support for the storage chip industry. Local governments have also established dedicated industrial funds, such as the hundred billion-level industrial funds in Shanghai and Hefei to support the development of the storage chip industry. In terms of policy support, the Chinese government has issued documents such as “Several Policies for Promoting High-Quality Development of the Integrated Circuit Industry and Software Industry in the New Era,” clearly supporting the R&D and capacity construction of storage chip technology. The Ministry of Industry and Information Technology, in conjunction with the National Development and Reform Commission, has launched the “Storage Chip Strengthening Project,” clearly requiring an increase in the self-sufficiency rate of DRAM to 25% and NAND Flash to 30% by 2025. In terms of technological breakthroughs, China has listed storage chips as a key area for national R&D plans, supporting companies in tackling key technologies. New storage technologies like RRAM have been included in the “bottleneck” technology breakthrough list and receive key support. At the same time, China is also reducing enterprise costs and improving competitiveness through tax incentives, R&D expense deductions, and accelerated equipment depreciation. In terms of industrial ecosystem construction, the Chinese government promotes the collaborative development of the storage chip industry chain, supporting companies in equipment, materials, design, manufacturing, and testing to progress together. By establishing industrial alliances and innovation centers, it promotes cooperation between industry, academia, and research, accelerating technological innovation and achievement transformation. 7.3 Security and Self-Control of the Industry Chain In the context of geopolitical tensions, the security and self-control of the storage chip industry chain have become a focus of attention for various countries. For China, achieving self-control in storage chips has significant strategic importance. China is the largest consumer market for storage chips globally, with annual consumption accounting for over 35% of the global total, but the self-sufficiency rate is only about 18%. Over-reliance on imports not only increases costs but also brings supply chain security risks. Therefore, accelerating the domestic production process of storage chips and improving self-control capabilities have become national strategies for China. China has made significant progress in self-control in storage chips. Yangtze Memory Technologies has achieved mass production of 232-layer 3D NAND, with technology levels reaching internationally advanced standards. Changxin Memory’s DDR5 yield has exceeded 80%, with technology catching up to the international mainstream. In terms of equipment and materials, Chinese companies are also accelerating domestic substitution, with Shanghai Micro Electronics’ lithography machines, Northern Huachuang’s etching machines, and Jiangfeng Electronics’ target materials being applied in domestic storage chip production lines. However, it should also be noted that China still faces many challenges in the storage chip field. In terms of technology, there is still a gap with international advanced levels, especially in high-end products and key technologies. In terms of the industrial ecosystem, a complete industrial chain has not yet formed, and some key equipment and materials still rely on imports. In terms of talent, the shortage of high-end technical talent remains a problem. For other countries, maintaining the stability and security of the storage chip industry chain is also crucial. Storage chip-producing countries like South Korea and Japan need to ensure their technological advantages and market positions. The U.S. maintains its dominant position in the storage chip field through technological blockades and industrial policies. Europe and Southeast Asia are also trying to attract investment through industrial policies to secure a place in the storage chip industry chain. Overall, the security and self-control of the storage chip industry chain have become a global focus, with various countries strengthening their competitiveness in this field through different means. This will drive profound changes in the global storage chip industry landscape and bring opportunities and challenges for the development of China’s storage chip industry. 8. Conclusion and Outlook Through an in-depth analysis of the storage chip industry chain, we can draw the following main conclusions: First, storage chips hold an irreplaceable strategic position in the digital economy era. As a core component of the semiconductor industry, storage chips are widely used in smartphones, personal computers, servers, data centers, game consoles, and various electronic devices, serving as a key infrastructure supporting the development of the digital economy. With the development of new technologies such as AI, 5G, and the Internet of Things, the demand for storage chips will continue to grow. Second, the global storage chip industry presents a highly concentrated oligopolistic structure. A few companies, including Samsung, SK Hynix, Micron, and Kioxia, occupy most of the global market share, holding absolute advantages in technology, production capacity, and market presence. Although Chinese companies started late, they are developing rapidly, with Yangtze Memory Technologies and Changxin Memory quickly catching up to international advanced levels. Third, storage chip production capacity is mainly concentrated in Asia, particularly in countries like China, South Korea, and Japan. China is not only the largest consumer market for storage chips globally but is also becoming an important production base. However, in terms of high-end capacity, South Korea and Japan still dominate. Fourth, the security and self-control of the storage chip industry chain have become a global focus. The U.S. attempts to curb the development of China’s storage chip industry through technological blockades and export controls, while China accelerates the domestic production process through policy support and financial investment. This will drive profound changes in the global storage chip industry landscape. Looking ahead, the storage chip industry will exhibit the following development trends: In terms of technological development, storage chips will continue to evolve towards higher density, larger capacity, lower power consumption, and higher performance. The number of stacked layers in 3D NAND will exceed 500 layers or even 1000 layers, HBM technology will achieve higher bandwidth, and new storage technologies like RRAM and MRAM will gradually become practical. At the same time, storage chips will deeply integrate with technologies like AI, cloud computing, and edge computing, driving changes in the entire IT architecture. In terms of market demand, AI and big data applications will become the most important driving force for the growth of storage chip demand, while emerging applications like autonomous driving and the Internet of Things will also bring significant market opportunities. It is expected that by 2030, the global storage chip market size will exceed $200 billion. In terms of industry structure, the global storage chip industry will shift from a highly concentrated structure to a multi-polar development. Chinese companies, supported by technological innovation and industrial policies, are expected to occupy a larger share of the mid-to-low-end market and gradually penetrate the high-end market. At the same time, other countries and regions will also attract investment through industrial policies, trying to secure a place in the storage chip industry chain. In terms of the industry chain, the storage chip industry chain will pay more attention to security and self-control. Countries will strengthen their layouts in key links such as equipment, materials, and design, promoting localization of the industry chain. China will accelerate the domestic production process of the entire storage chip industry chain, improving self-control capabilities. Recommendations for industry participants: For Chinese storage chip companies, they should adhere to technological innovation and industrial upgrading, strengthen cooperation with upstream and downstream companies in the equipment and materials sectors, and jointly promote industry development. At the same time, they should focus on talent cultivation and technological accumulation to gradually narrow the gap with international advanced levels. For investors, the storage chip industry has characteristics of high investment, high risk, and high returns, requiring a long-term investment perspective and risk tolerance. It is recommended to focus on companies with core technologies, complete industry chain layouts, and policy support. For policymakers, they should continue to increase support for the storage chip industry, especially in key technology breakthroughs, industrial ecosystem construction, and talent cultivation. At the same time, they should strengthen international cooperation to enhance the competitiveness of the industry in an open environment. In summary, the storage chip industry is at a critical period of technological transformation and structural reshaping, with both opportunities and challenges. Only by adhering to innovation-driven, open cooperation, and security and controllability can we occupy a favorable position in the fierce international competition and promote the healthy and sustainable development of the industry.