Murata Manufacturing halts general MLCC orders, 20-week lead times

The passive component industry is experiencing a structural shift driven by AI compute demand. MLCCs, long considered a commodity, have become a strategic resource as AI servers require over 10 times more units than general servers. Murata's decision to refuse general-purpose orders reveals that capacity expansion cannot keep pace, with even its 800億円 investment yielding only a 10–15% increase by fiscal 2027. This forces a prioritization of high-margin AI components over legacy products, sending ripple effects through the supply chain.

Competitors are responding aggressively to fill the gap. Samsung Electro-Mechanics secured its largest single MLCC long-term contract at 1兆722億ウォン, while Taiwanese makers like Yageo and 華新科 are absorbing spillover demand. The market is clearly supplier-driven, with inventory at record lows and capacity utilization at 95%. This tightness is not merely cyclical; it reflects a fundamental mismatch between AI's exponential demand and the industry's incremental capacity additions.

Looking ahead, the physical limits of traditional MLCCs in high-frequency applications are opening the door for silicon capacitors. These devices, manufactured using semiconductor wafer processes, offer lower parasitic parameters and better integration with advanced packaging. However, their custom nature — requiring tight alignment with chip architectures and packaging processes — remains a hurdle. The outcome likely depends on whether silicon capacitors can scale beyond niche applications to become a standard option, which would test the dominance of the Japanese ceramic powder supply chain and alter the competitive landscape for passive components.