Soochow Securities Company Limited released a research report highlighting that the rising power density and supply current demands of AI servers are significantly increasing the value of AI chip inductors. The rapid escalation of single-GPU power consumption is driving both volume and price growth for AI chip inductors across entire racks. The global market size for these components has expanded from RMB 1.1 billion in 2020 to RMB 5.8 billion in 2025, with projections indicating it will surge to RMB 26.9 billion by 2030. Domestic players are quickly catching up in this emerging sector, and TLVR technology is poised to drastically boost inductor value. In terms of investment recommendations, the firm emphasizes that higher AI server power density and supply current directly amplify the value proposition of AI chip inductors.
Inductors: Essential passive components for current stabilization
AI chip inductors primarily serve third-stage power conversion scenarios. Inductors rank as one of the three fundamental passive components, second only to capacitors in market size. They are categorized by application into traditional inductors used in consumer electronics, automotive, and photovoltaic sectors, and AI chip inductors designed for third-stage power delivery. In data center environments, traditional wound inductors handle first and second-stage conversions for server room and motherboard power supply, whereas AI chip inductors support third-stage circuits powering high-performance accelerator chips such as GPUs, ASICs, and FPGAs. These inductors also pair with optical modules and high-speed memory chips, and can meet the power supply needs of edge-side AI hardware.
The rapid increase in single-GPU power consumption is fueling simultaneous volume and price gains for AI chip inductors across complete systems. Nvidia's single-GPU power draw has now climbed to approximately 1800-2300W for the current Rubin series. This rise in supply current demands a greater number of power phases, with a gradual transition toward VPD architecture, and each phase requires one dedicated chip inductor. Concurrently, higher current density and thermal management requirements are pushing AI chip inductors toward higher rated currents, superior high-frequency performance, and enhanced reliability. These demands inherently drive advancements in upstream alloy magnetic powder quality and downstream processes like TLVR and copper-iron co-firing. Future TLVR inductors could command prices 3-5 times higher than conventional ones, propelling per-rack chip inductor value upward. Estimates suggest the new-generation VR200 NVL72 rack could house 6000-7000 chip inductors, with corresponding value reaching tens of thousands of RMB.
Driven by surging computational demands, high-performance chips, and next-generation memory deployment, the global AI chip inductor market has grown from RMB 1.1 billion in 2020 to RMB 5.8 billion in 2025, reflecting a CAGR of 38.2%. It is expected to reach RMB 26.9 billion by 2030, with a projected CAGR of 36.0%.
Established players lead traditional segment; domestic firms race ahead in emerging AI sector
The traditional inductor market is fragmented with limited growth, dominated by Japanese and Taiwanese manufacturers such as TDK, Murata, and Chilisin. In contrast, AI chip inductors offer gross margins 10-30 percentage points higher than traditional types, making them highly profitable. Against the backdrop of growing AI server demand, domestic companies are swiftly entering the AI chip inductor space, with Sunlord and Boke already ranking among the global top five. The short product life cycles and rapid downstream iteration of chip inductors are prompting Chinese manufacturers to accelerate investment in high-end production capacity for TLVR and chip inductors, aiming to secure positions in major supply chains. As capacity expands and competition intensifies, prices for mainstream hot-pressed integrated molding inductors are under pressure. However, the increased technical complexity of next-generation TLVR inductors is expected to lift unit prices, providing additional value growth.
Magnetic materials and customer relationships form key competitive barriers
Magnetic materials currently account for 50-70% of inductor raw material costs. High-end magnetic powders such as iron-nickel and amorphous alloys offer broad application potential but present significant technical challenges and carry high prices. Magnetic powder formulations can lower downstream inductor processing requirements, bypassing yield issues associated with novel technologies like copper-iron co-firing. In the manufacturing segment, customer certification cycles are lengthy, fostering strong client loyalty. AI server makers require highly customized part numbers, necessitating close collaboration with suppliers through testing, validation, and pre-research phases.
Risk factors
Risks include a slowdown in AI capital expenditure growth, substitution by alternative technologies, intensifying price competition, fluctuations in raw material costs, geopolitical tensions and trade frictions, slower-than-expected production ramp-ups, and customer concentration risks.