Opportunities for MLCC under the AI Wave: A Comprehensive Analysis of Core A-Share Stocks

Under the AI wave, the demand for MLCC in AI servers (especially those equipped with high-performance GPU/TPU computing clusters) is showing a "dual increase" in both quantity and quality, resulting in an explosive growth trend. A typical AI server requires approximately 15k to 25k MLCCs, with very high requirements for high frequency, high capacitance, high voltage, and high-temperature resistance.
In the A-share market, although high-end AI-grade MLCCs are still dominated by Japanese and Korean giants (such as Murata and Samsung Electro-Mechanics), several domestic leading companies have achieved deep penetration in the AI supply chain, upstream core materials, and specialized/military-grade AI chip supporting fields.
1. Leading MLCC Manufacturers: Direct Beneficiaries of AI Computing Power Expansion
These two are the absolute mainstays of A-share MLCCs, accelerating breakthroughs in high-end, high-capacitance, high-voltage MLCC technology required for AI servers.
Sanhuan Group: A leader in domestic MLCC capacity and technology tiers. In response to the high power demands of AI servers and data centers, the company has successfully launched high-voltage, high-capacitance series products such as 0805-105-100V, addressing power management pain points around high-performance chips. Its market share in network communication and servers is steadily increasing. With an integrated advantage of "materials + devices," it is one of the most capable A-share companies to undertake domestic AI server MLCC orders.
Fenghua High-Tech: As one of the largest passive component manufacturers in China, it has a deep layout in ultra-micro MLCCs (such as 01005, 0201 specifications) and high-frequency, high-reliability MLCCs required for AI servers. As industry prosperity improves due to AI, its capacity utilization remains high, making it an important beneficiary of domestic AI hardware supply chain localization procurement.
2. Upstream Core Material Leaders: The Unsung Heroes of AI-grade MLCCs
MLCCs used in AI servers are extremely miniaturized and layered with many sheets, demanding nearly stringent requirements for upstream ceramic powders and electrode metal powders. These two companies have high barriers and significant influence in the industry chain.
Guoci Materials: A core supplier of ceramic powders for global MLCCs. High-end AI-grade MLCCs must use ultra-fine, high-purity barium titanate formulation powders. Guoci Materials is one of the few companies worldwide that master this technology, holding a leading global market share. Whether domestic companies like Sanhuan and Fenghua or some overseas giants, their production of high-end MLCCs relies on Guoci’s formulations. The explosive demand for high-end MLCCs driven by AI will directly boost Guoci’s powder sales.
Boxin New Materials: A global leader in ultra-fine electronic-grade nickel powders. The internal electrodes of MLCCs require extremely fine nickel and copper powders. AI server MLCCs tend toward "multi-layering"—more layers mean higher capacitance—which causes the consumption of high-quality nickel powders to increase exponentially. As a core supplier, Boxin New Materials’ products are highly aligned with the upgrade pace of AI hardware.
3. RF and Special-Grade MLCCs: AI Edge Computing and High-Reliability Computing Power
Besides general-purpose servers, AI also requires special MLCCs for industrial control, edge computing, and high-reliability computing platforms.
Dalikai Pu: Focused on RF microwave MLCCs, mainly used for high-frequency signal transmission. In the AI era, with upgrades in data center optical modules (such as 800G/1.6T optical modules) and high-speed communication base stations, demand for high-frequency, low-loss RF MLCCs has surged. Dalikai Pu has a strong domestic substitution advantage in this niche, serving as an invisible champion for optical modules and high-speed communication supporting.
4. Investment Logic Summary
The layout of AI-related MLCC concepts in A-shares can be viewed through three layers.
The first layer is device substitution, emphasizing technological breakthroughs and volume growth of domestic orders, with Sanhuan Group as the top choice. Its "materials + device" integrated cost structure makes it the strongest candidate for domestic substitution of high-end MLCCs in AI servers.
The second layer is material barriers, focusing on industry moat and stable monopoly profits. Guoci Materials and Boxin New Materials, with their irreplaceable positions in powder materials, have stronger risk resistance. Regardless of who wins downstream, demand for upstream materials will continue to grow.
The third layer is niche segments, emphasizing high-frequency communication and optical modules, which are independent growth drivers outside the main server line. Dalikai Pu’s RF MLCCs have growth potential amid large-scale deployment of 800G optical modules.
Overall, MLCCs are the passive components most easily overlooked in AI computing infrastructure but have the most certain demand. Quantity growth comes from the expansion of server units, while quality upgrades come from the increased value per server. The combination of these factors indicates that this supply chain’s prosperity cycle has only just begun.