Minsheng Securities: AI-driven intelligent computing demand highlights SST solutions leading the transformation of data centers

According to Zhitong Finance APP, Minsheng Securities released a research report stating that the global AI market is expected to have a compound annual growth rate (CAGR) of over 35% from 2023 to 2030, surging from 1.1879 trillion yuan to 11.4554 trillion yuan. The explosive growth of the AI industry is driving an upgrade in demand on the intelligent computing side, while data center supply and distribution technologies face challenges of high density. The firm focuses on recommending the SST (Solid State Transformer) solution industry chain, specifically paying attention to: 1) traditional enterprises with technological accumulation in the high-frequency transformer field; 2) companies with power electronics three-level circuit topology technology; 3) upstream material suppliers such as nanocrystalline/ferromagnetic cores. The main viewpoints of Minsheng Securities are as follows: The global AI market size CAGR is expected to exceed 35% from 2023 to 2030, highlighting AI-driven demand on the intelligent computing side. The investment wave in the global AI market has emerged. According to Precedence Research market forecasts, the global artificial intelligence (AI) market is expected to grow from 1.1879 trillion yuan to 11.4554 trillion yuan from 2023 to 2030, achieving a CAGR of over 35%. With the rapid development of cloud computing, more and more critical applications are migrating to the cloud, leading to larger scales of data center construction, and the requirements for latency, bandwidth, availability, and security in computing centers have become higher. HVDC, Panama power supply, and other solutions have certain advantages, with the SST solution standing out in comprehensive advantages. The evolution of rack power towards high density poses several challenges for data center supply and distribution architecture, including increased demand for all-DC systems, higher terminal voltage levels, and increased footprint for supply and distribution. The HVDC solution has certain upgrades compared to traditional UPS solutions, and the Panama power supply solution further upgrades to achieve modularity and high efficiency. The SST solution is suitable for DC power scenarios in microgrids, capable of directly converting AC power to 800V DC supply, while also being lightweight, autonomously solving harmonic management issues, quickly detecting and isolating input or output side faults, and supporting bidirectional power flow, which has been applied in data center projects. Investment recommendations for data center SST solutions High-frequency transformers: The core component of the SST solution is the high-frequency transformer, and traditional transformer companies have deep technological accumulation. It is recommended to pay attention to: leading special transformer company Xinte Electric, leading ultra-high voltage transformer company & parent company of Xi'an Electric Power Electronics China Xi'an Electric, and leading AIDC transformer enterprise Jinpan Smart Technology. Power electronic devices: SST achieves power conversion and isolation through three-level circuit topology, and companies with relevant technological reserves in power electronics are expected to enjoy the benefits of the sector. It is recommended to pay attention to: multi-product layout & established power electronics company Shenghong Co., Ltd., series solution provider for power electronics Sifang Co., Ltd., and companies with deep layouts in rectification & inversion technology such as szclou. Upstream raw materials: High-frequency transformers use nanocrystalline/ferromagnetic cores. It is recommended to pay attention to: leading advanced magnetic metal material company Yunlu Co., Ltd., and leading company in the field of amorphous materials AT&M. Risk Warning: Downstream capital expenditure may be less than expected; risk of intensified industry competition; risk of technological substitution