Orient Securities: Metal Soft Magnetic Powder Cores Essential for ASIC Power Efficiency, Future Inductor Material Usage Expected to Grow Further

Stock News
10/10

Orient Securities has released a research report stating that due to the unavoidable PDN voltage drop and losses in traditional horizontal power supply modes, as well as the synchronous growth of PDN losses driven by increasing AI computing power consumption, the core pain point for end users after deploying ASICs is power saving. Metal soft magnetic powder cores are essential for ASIC power efficiency. As DDR6 standards continue to upgrade memory performance, energy efficiency, and power supply solutions compared to previous generations, inductor material usage is expected to grow further. With overseas AI giants accelerating large-scale deployment of self-developed ASIC chips next year, the application proportion of chip inductors in AI servers may increase significantly.

Orient Securities' main viewpoints are as follows:

**As overseas AI giants shift to self-developed ASIC chips, power saving has become the core pain point for end users**

With the development of AI computing power, AI computing chips represented by NVIDIA's GB300 have reached new heights in parameters and performance, while power consumption has also increased significantly. Since ASICs (Application-Specific Integrated Circuits) customize hardware architecture for single tasks, which highly matches AI inference requirements, and have obvious advantages over GPUs in multiple aspects including area, energy consumption, integration, and price, and significantly outperform GPUs in the core metric of computing power-to-power consumption ratio, overseas AI giants are turning to self-developed ASIC chips. Among them, Meta, through cooperation with Broadcom, plans to launch its first AI ASIC chip MTIAT-V1 as early as the fourth quarter of this year.

Due to the unavoidable PDN voltage drop and losses in traditional horizontal power supply modes, as well as the synchronous growth of PDN losses brought by increasing AI computing power consumption, the core pain point for end users after deploying ASICs is power saving.

**Metal soft magnetic powder cores are essential for ASIC power saving**

ASIC power-saving requirements place higher demands on the structure and materials of power modules and inductors. In terms of structure, power modules adopting vertical stacking design break the PCB board size limitations of planar layouts, achieving PDN path reduction to millimeter level through close coupling of power modules and processors, significantly reducing PDN losses. In terms of materials, due to the miniaturization of power module structures and high-power, high-current application scenarios in AI servers, higher requirements are placed on the high-current carrying capacity of inductors.

Molded inductors using metal soft magnetic powder cores have higher saturation magnetic flux density compared to ferrite, enabling them to withstand larger currents in unit volume and possess better high-temperature stability, while having lower core losses and simpler molding processes, making them indispensable new inductor materials for ASIC power saving.

**ASIC volume growth in 2026 and medium-term DDR6 memory power management bring broad growth space for metal soft magnetic powder**

Currently, overseas AI giants are competing to advance self-developed ASIC chip layouts. Meta has signed long-term capacity agreements with TSMC, planning to launch multiple MTIA series chips in phases between late 2025 and 2026. Microsoft plans to mass-ship self-developed ASIC chips in 2027, with inductor demand expected to see significant growth. NVIDIA AI GPU shipments are expected to maintain growth, and AI computing chips may enter a GPU+ASIC dual-drive era in 2026, with total inductor demand potentially exceeding 400 million pieces.

Regarding DDR memory, high-capacity and high-speed DDR5 particles place higher requirements on power integrity. DDR5 uses PMIC power supply solutions for the first time, and due to changes in power supply architecture, inductor forms have also evolved from large-size, high-current assembled types to small-size, thin inductors, similarly placing higher requirements on power modules and inductor materials. As DDR6 standards continue to upgrade memory performance, energy efficiency, and power supply solutions compared to previous generations, inductor material usage is expected to grow further.

**Industry leader in metal soft magnetic powder cores has technological advantages, with medium-term capacity expected to grow rapidly**

The company is a pioneer in the industrialization of gas atomization powder production, mastering gas atomization metal powder preparation technology that has advantages over water atomization powder including higher sphericity and lower oxygen content suitable for pressing into powder cores. The company's conventional and ultrafine metal alloy soft magnetic powders have finer particle sizes and lower oxygen content compared to peer products, providing stronger product competitive advantages.

In terms of capacity, the company's fundraising projects are expected to be fully completed by the end of 2025, with capacity ramp-up beginning in 2026, potentially reaching total capacity of 300 million pieces. With production line efficiency improvements, the company's inductor material capacity may have further growth potential in the medium term.

**Risk Warnings**

Downstream industry demand for inductors falling short of expectations, intensified industry competition leading to declining profitability, capacity deployment progress falling short of expectations, US tariff policy change risks, changes in assumption conditions affecting calculation results.

免责声明:投资有风险,本文并非投资建议,以上内容不应被视为任何金融产品的购买或出售要约、建议或邀请,作者或其他用户的任何相关讨论、评论或帖子也不应被视为此类内容。本文仅供一般参考,不考虑您的个人投资目标、财务状况或需求。TTM对信息的准确性和完整性不承担任何责任或保证,投资者应自行研究并在投资前寻求专业建议。

热议股票

  1. 1
     
     
     
     
  2. 2
     
     
     
     
  3. 3
     
     
     
     
  4. 4
     
     
     
     
  5. 5
     
     
     
     
  6. 6
     
     
     
     
  7. 7
     
     
     
     
  8. 8
     
     
     
     
  9. 9
     
     
     
     
  10. 10