AI servers drive MLCC into a structural upcycle

Overview

This report discusses the structural impact of AI server and data-center buildout on the MLCC industry. The core view is that AI's requirements for power integrity, low ESR, low ESL, high capacitance, and small form factors are significantly raising MLCC demand, shifting it from volume expansion toward high-end products, specification upgrades, and long-term supply lock-ins. The report argues that this cycle differs from a traditional commodity cycle and is closer to the structural tightness created by AI-driven capacity shifts in DDR4 and HBM.

Core views

AI servers are causing a step-change increase in both MLCC unit consumption and value content, while high-end MLCC capacity is not easily interchangeable with consumer, automotive, or industrial specifications. As leading manufacturers redirect capacity toward AI and other high-margin applications, the advanced capacity available to traditional end markets is being squeezed. Because new capacity takes a long time to build, equipment is highly customized, and customer qualification is strict, supply elasticity is limited, so pricing and margin improvement may continue into 2H26 and 2027.

Analysis framework

The report combines channel checks, supply-demand estimates, product-specification evolution, competitive landscape analysis, company earnings revisions, and a risk-reward valuation framework to assess the impact of AI on total MLCC demand, product mix, capacity consumption, pricing, margins, and related stock performance.

Methodology notes

• Supply and demand analysisAI server MLCC content and capacity consumption estimate

  Usage, value, specifications, and capacity intensity jointly determine the supply-demand gap

  The report compares MLCC usage and value content across traditional servers, HGX, GB300, VR200, and other platforms, and emphasizes that high-capacitance AI MLCCs require more dielectric layers, so each unit consumes more capacity than ordinary products.

• Channel researchPrice, inventory, and order trend checks

  Spot prices, contract prices, months of inventory, and book-to-bill ratios reflect cycle position

  The report cites spot prices in Shenzhen Huaqiangbei, distributor pricing, direct-sales pricing, inventory levels, and long-term supply agreement negotiations to judge whether supply tightness is shifting from short-term speculation to a more durable trend.

• Relative competitive analysisJapan-Korea-Taiwan MLCC peers comparison

  Profitability, high-end exposure, growth, cash conversion, and ROIC are used to compare vendors

  The report compares the market share, margins, growth outlook, cash flow, and capital returns of Murata, Samsung Electro-Mechanics, Yageo, Taiyo Yuden, TDK, and others.

• Valuation methodsRisk Reward / Residual Income Model

  Target price is derived from earnings revisions and shareholder equity value

  The report raises Samsung Electro-Mechanics' target price to W2,560,000 and implies about 20% upside, driven mainly by MLCC earnings upgrades and contributions from AI-related businesses.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Samsung Electro-Mechanics (009150.KS)

  Core beneficiary; the report raises earnings estimates and target price

  Strengths

  AI MLCCs, ABF substrates, embedded MLCCs, silicon capacitors, and potential glass substrates support growth; the Tianjin plant is running at full capacity; valuation remains attractive relative to historical averages.

  Weaknesses

  It still needs to continue scaling high-end capacity, yields, and AI customer commitments, and the share price has already risen sharply.

  Comparison

  Still catching up with Murata, but the gap is narrowing; the report believes the roughly 20% target-price upside supports the recommendation.

  Risks

  AI demand falls short of expectations, prices retreat, expansion or yields disappoint, or valuation de-rates.

• Murata Manufacturing (6981.T)

  Leader in high-end AI MLCCs and top pick in Japan

  Strengths

  AI server orders are about 2x capacity, with utilization at 90%-95%; it leads in materials vertical integration, ultra-small 008004 formats, and high-end market share.

  Weaknesses

  Capacity tightness may limit short-term shipment flexibility, and market expectations may already be high.

  Comparison

  The report believes Murata may maintain a clear leadership position, with about 41% MLCC market share in 2025.

  Risks

  Pace of expansion, customer platform shifts, price regulation, or intensifying competition.

• Yageo Corp. (2327.TW)

  Preferred Taiwan name, benefiting from spillover effects as capacity shifts from standard MLCCs to high-end products

  Strengths

  Business transformation, price increases, and record backlog; 1Q26 revenue and gross margin were better than expected.

  Weaknesses

  M&A contribution remains uncertain, and standard-product utilization is still below high-end products.

  Comparison

  Compared with Japanese and Korean leaders, it has a smaller direct share of high-end AI, but benefits from improved pricing in commodity and specialty products.

  Risks

  M&A integration, end-demand volatility, and weaker-than-expected price pass-through.

• Taiyo Yuden (6976.T)

  Expected to underperform relative

  Strengths

  MLCC makes up a high share of sales, and AI/DC-related MLCC sales are expected to increase.

  Weaknesses

  Capacity expansion is constrained, with F3/26 capacity rising only about 5%; some production issues are affecting utilization; market expectations are high.

  Comparison

  Relative to Murata and SEMCO, the earnings leverage from AI upside may be smaller.

  Risks

  If AI demand accelerates further and the company improves utilization, performance could surprise to the upside.

• TDK, CCTC, Walsin Tech, Guangdong Fenghua Advanced Technology, Chaozhou Three-Circle, Eyang Technology

  Other competitors or supply-chain participants

  Strengths

  Some companies focus on high-voltage, industrial modules, or domestic server supply chains, and may benefit from tight supply and domestic substitution.

  Weaknesses

  Most commodity-grade players mainly serve consumer electronics, and barriers to high-end AI qualification and materials/processes are high.

  Comparison

  Weaker position in the high-end AI MLCC market than Murata and Samsung Electro-Mechanics.

  Risks

  Slow high-end entry, price-cycle reversal, or customer qualification failure.

Key data

• Samsung Electro-Mechanics target priceW2,560,000Raised from W920,000 to W2,560,000.
• AI server MLCC usage~440,000 units per AI serverAbout 10-15x that of a traditional server.
• AI-driven MLCC demandMore than 4x growth from 2025-2030The report expects AI-related demand to outgrow the industry total by a wide margin.
• Industry capacity growth10%-15% annuallyMajor players are avoiding speculative capacity expansion, and new capacity takes a long time to build.
• Global MLCC shipment valueUS$14.7bn in 2025 to US$24.3bn in 2028eImplying about 18% CAGR, above the 10-year trend of 6.5%.
• AI server MLCC TAM~US$900M by 2027e; could exceed US$1bn by 2030eMaterially above the prior estimate of about US$550M for 2030.
• High-end AI MLCC market shareMurata 45%, Samsung Electro-Mechanics 40%The Japanese-Korean duopoly together accounts for about 85%.
• Channel price changesdistributor prices +200%-300% since 1Q26Some spot specifications have risen 2-10x, but the report distinguishes end-demand from trader hoarding.
• Inventory levelsdistributors ~1.5 months; downstream customers <1 monthAfter two years of destocking, supply-chain inventories are low.
• High-end lead times>20 weeksShortages in high-end products remain real.

Impact & implications

If the report's view is correct, the MLCC industry's earnings driver will shift from simple shipment recovery to a higher share of high-end products, AI customer lock-ins, rising prices, and capacity scarcity. Samsung Electro-Mechanics, Murata, and Yageo benefit from AI and high-end exposure; standard-product customers, consumer electronics, and some automotive and industrial applications may face higher procurement costs and lower supply priority.

Risks

• AI server or data-center demand falls short, weakening high-end MLCC orders and pricing momentum.
• Channel spot-price increases include trader hoarding; if end demand fails to catch up, short-term prices could retreat.
• Faster-than-expected capacity additions, yield improvements, or substitution technologies from leading players could ease supply tightness.
• MLCC-related stocks have risen sharply year to date, and valuations already reflect some optimism.
• If consumer, automotive, and industrial demand soften, they could offset part of the AI-driven increment.

What to watch

• AI server specification evolution and changes in MLCC usage on next-generation platforms.
• Whether the book-to-bill ratio shows a sharp reversal.
• YoY MLCC price momentum, especially contract prices rather than only spot prices.
• AI server and data-center capex demand.
• High-end MLCC lead times, inventory months, and long-term supply agreement signings.
• Capacity utilization, gross margin, and AI-related revenue mix at Murata, Samsung Electro-Mechanics, Yageo, and Taiyo Yuden.