Shanghai Securities: Enterprise-level storage is expected to remain highly prosperous, domestic enterprise-level SSD-related manufacturers are expected to benefit from the layout.

Shanghai Securities released a research report stating that IDC predicts that the market size of China's enterprise solid-state drive (SSD) market will reach 9.1 billion US dollars by 2029. With the increase in capacity and decrease in unit cost of enterprise SSDs, there is potential to accelerate the replacement of HDDs. Policy support is driving the increase in domestic production rate, with local storage companies benefiting from technological innovation in large capacity products, controller chips, and localization of the industry chain and services, narrowing the gap with international brands, and expected to increase overall market share. The company believes that the enterprise storage sector is expected to remain prosperous, and domestic companies positioning in enterprise SSDs are expected to benefit. Shanghai Securities' main points are as follows: - Enterprise SSDs are applied in high-performance computing, data centers, and other scenarios. - According to the Forward Industry Research Institute, enterprise solid-state drives (eSSD) are used in various enterprise scenarios such as high-performance computing, edge computing, high-end storage, and data centers. They have uninterrupted working capabilities, and can handle I/O-intensive workloads such as database files, index logs, data analysis, and other high-performance transaction processing. Compared to consumer-grade SSDs, enterprise SSDs target enterprise-level users, require data security, stable speed, and long-term durability, with higher reliability and performance, and endurance. IDC data shows that the market size of China's enterprise SSDs reached 6.25 billion US dollars in 2024, with a year-on-year growth of 187.9%. IDC predicts that by 2029, the market size of China's enterprise SSDs will reach 9.1 billion US dollars. - HDD is facing a supply-demand gap, with enterprise SSDs expected to accelerate the replacement of HDDs. - Compared to HDDs, SSDs have superior reading and writing speeds due to the elimination of mechanical components and the elimination of rotational and seek delays. Enterprise QLC SSDs are the mainstream replacement solution for enterprise-level HDDs due to their large capacity and low energy consumption advantages. HDDs still dominate in conventional large-scale cold data or warm data storage scenarios due to cost-effectiveness. However, due to the lack of expansion plans by major HDD manufacturers in recent years, they are unable to meet the sudden and massive storage demands stimulated by AI. According to TrendForce, the delivery lead time for NLHDD has rapidly extended to over 52 weeks from several weeks. The North American market is facing a shortage of HDDs for data centers, leading to some North American customers shifting their purchasing demands to enterprise storage. Additionally, with the increase in capacity and decrease in unit cost of enterprise SSDs, there is potential to accelerate the replacement of HDDs. - The value of enterprise SSDs in AI servers is higher than in traditional servers, with the capacity of SSDs in AI inference stage expected to increase. - In the AI era, large models in training and inference processes have higher requirements for data transmission, operations, and storage. The stable and rapid growth of the AI server market will drive downstream market demand for large-capacity QLC SSDs, boosting the growth of enterprise SSDs. The value of enterprise SSDs in a single AI server is generally more than three times that of a general-purpose server. According to IDC forecast, the global AI server market size will reach 34.71 billion US dollars by 2026, with a compound annual growth rate of 17.3% from 2021 to 2026. Furthermore, the capacity of SSDs in the AI inference stage is expected to increase compared to the training stage. In the training stage, SSDs are responsible for storing model parameters, including continuously updating weights and biases, and are also used to create checkpoints to periodically save AI model training progress to ensure recovery even if training is interrupted at a specific point. These functions heavily rely on SSDs' high-speed transmission and write durability. Therefore, customers mainly choose to use 4TB and 8TB TLC (Triple-Layer Cell) SSD products to meet the demanding training process requirements. In the AI inference stage, as more generated information is displayed in video or image form, the data storage volume also increases accordingly. Therefore, large capacity SSDs such as 16TB and above TLC/QLC (Quad-Layer Cell) become the main products used for AI inference. - Industry landscape: Solidigm and Samsung are global leaders, with domestic manufacturers having significant room for improvement in market share. - According to IDC data, from the perspective of manufacturer sales, Solidigm and Samsung are the global leaders in the enterprise SSD market in China, holding the largest market share. Samsung maintains its global leadership position with its full-stack technology capabilities and high-end products, but faces domestic substitution and cost pressures in the Chinese market. Solidigm focuses on AI-optimized storage and core large-capacity products, relying on SK Hynix's rapid rise in the ec...