--- title: "The Next Flashpoint in the AI Arms Race! Goldman Sachs: Capacitors Are the \"New Memory\"" type: "News" locale: "en" url: "https://longbridge.com/en/news/288144670.md" description: "Goldman Sachs states that Multilayer Ceramic Capacitors (MLCCs) are becoming the next core bottleneck in AI infrastructure. Driven by a fourfold surge in demand from AI servers and Rubin racks, lead times for high-end MLCCs have exceeded 20 weeks, prompting recent price hikes by Japanese giants. In April, Japan's MLCC export value surged 28% year-on-year, ushering the industry into a long-term cycle of rising volume and prices, with leading companies poised to release significant earnings elasticity" datetime: "2026-05-30T12:00:53.000Z" locales: - [zh-CN](https://longbridge.com/zh-CN/news/288144670.md) - [en](https://longbridge.com/en/news/288144670.md) - [zh-HK](https://longbridge.com/zh-HK/news/288144670.md) --- # The Next Flashpoint in the AI Arms Race! Goldman Sachs: Capacitors Are the "New Memory" In the infrastructure arms race of the AI supercycle, the rotating bottlenecks in supply and demand have spawned wave after wave of market winners—data centers, energy infrastructure, and memory chips have successively become the focus of capital pursuit. Goldman Sachs' latest analysis points out that **the next bottleneck poised to explode is the long-overlooked Multilayer Ceramic Capacitor (MLCC), which it characterizes as the "New Memory"**—a niche segment of passive components standing at the starting point of a cycle featuring both rising volumes and prices. Price signals are rapidly strengthening. According to media reports in April, Murata Manufacturing raised prices for its MLCC products used in AI servers and high-end automotive applications by 15% to 35% effective April 1; Taiyo Yuden also notified customers that it would implement price increases for certain product lines starting in May, citing the continuous rise in costs of various raw materials, including precious metals. Spot prices for AI-related MLCCs have risen by approximately 20%, with lead times for high-end products exceeding 20 weeks. Trade statistics released by Japan's Ministry of Finance on May 28 further corroborate this trend: the average export price of MLCCs in April rose 16% year-on-year, while the export value increased by 28%. Goldman Sachs analyst Daiki Takayama expects the MLCC market size for AI servers to grow more than fourfold from FY2025 to FY2030, expanding from approximately JPY 215 billion to around JPY 920 billion, representing a compound annual growth rate (CAGR) of 34%. Meanwhile, **the annual capacity growth rate for the entire MLCC industry is only slightly above 10%, with expansion constrained by internal self-production capabilities for equipment and materials.** Goldman Sachs stated that the current AI-driven MLCC cycle "will be the largest and longest-lasting in history, and we believe it is still in its early stages," maintaining Buy ratings for Murata Manufacturing, Taiyo Yuden, and TDK Corporation. For investors, the earnings elasticity brought about by the MLCC supply-demand mismatch cannot be underestimated: a mere 5% increase in average selling prices could theoretically boost Murata Manufacturing's operating profit in FY2027 by approximately 13%, and Taiyo Yuden's operating profit by as much as 37%. Goldman Sachs' Asian MLCC thematic stock basket has recently begun to strengthen, but there remains significant catch-up potential compared to other popular AI themes. ## The Role of Capacitors: "Instant Voltage Regulators" Beside AI Chips MLCCs can be understood as extremely miniature, ultra-high-speed charge-discharge units. While ordinary batteries store large amounts of electrical energy and release it slowly, MLCCs store small amounts of energy but can complete charge-discharge cycles within fractions of a millisecond. Their core functions are twofold: first, power smoothing—absorbing instantaneous voltage spikes or filling sudden drops to provide stable current to sensitive chips; second, noise filtering—blocking electrical interference that could corrupt digital data. The operational characteristics of AI servers make MLCCs indispensable. When AI models process large-scale computations, processors instantly spike power consumption at the microsecond level, only to drop sharply to zero once the computation ends; power systems simply cannot respond quickly enough to such changes. MLCCs are installed directly adjacent to AI chips, releasing energy instantly during peak demand to prevent server crashes. Since AI chips (such as NVIDIA GPUs) need to handle billions of tasks simultaneously, a top-tier AI server rack may require up to 600,000 MLCCs working in concert to maintain system stability. In the view of Goldman Sachs analyst Nelson Armbrust, **MLCCs have become the third-highest cost component in the Bill of Materials (BOM) for AI servers, trailing only GPUs and memory.** The total MLCC market size is approximately $15 billion, with the server sub-segment accounting for about $1.3 billion and expanding at an 80% CAGR, while demand growth in other application areas such as automobiles and mobile phones has slowed. Daiki Takayama expects the share of MLCCs in the AI server BOM to rise from the current ~0.5% to approximately 1%. ## Supply-Demand Imbalance: 10% Capacity Growth Facing a 4x Demand Shock Goldman Sachs analyst Allen Chang clearly pointed out the core structural contradiction in the MLCC industry: annual capacity growth is only slightly above 10%, and since both equipment and materials rely on internal production, expansion progress is limited by internal engineering resources, making it difficult to accelerate significantly. The demand shock from the AI server sector, however, is on an entirely different scale. **Goldman Sachs estimates that demand for MLCCs from AI servers will grow approximately 4.3-fold from FY2025 to FY2030.** At the same time, demand for high-voltage, high-capacity MLCCs driven by automotive electrification remains strong—the MLCC content per electric vehicle continues to rise. These two demand pillars are jointly absorbing the already limited new capacity, causing customers in the consumer electronics sector, despite facing their own declining demand, to actively seek long-term supply contracts to guard against future supply shortage risks. Signals of tight market supply and demand are evident on multiple levels: lead times for high-end MLCCs (high-capacity, high-voltage specifications) have exceeded 20 weeks; spot and distributor channel prices for low-capacity and consumer-grade MLCCs have risen 20% to 40% due to hoarding and double-ordering; high prices for key raw materials like nickel and silver are creating cost pressure across all sub-categories; although OEM contract prices have not yet been raised significantly, market expectations are shifting. From the perspective of the overall AI supply chain timeline, Goldman Sachs' analytical framework shows that the price start for MLCCs lagged significantly behind other core AI components such as DRAM, NAND memory, ABF substrates, and Copper Clad Laminates (CCL). Precisely because of this, **Goldman Sachs judges that MLCCs (alongside ABF and CCL) have the longest price upside potential among all AI components and materials.** ## Accelerating Price Hike Cycle: Dual Validation from Data and Earnings Elasticity The price hike measures by Murata Manufacturing and Taiyo Yuden mark the official start of the MLCC price cycle. Murata Manufacturing raised prices by 15% to 35% for AI server and high-end automotive applications effective April 1; Taiyo Yuden notified customers that it would implement price adjustments for multiple product lines starting in May, including MLCCs, inductors, RF devices, FBAR/SAW devices, and aluminum electrolytic capacitors. Trade statistics released by Japan's Ministry of Finance on May 28 provided macro validation: in April, the average export price of MLCCs rose 3% month-on-month and 16% year-on-year; export volume increased 10% year-on-year; and export value grew 28% year-on-year. Goldman Sachs believes these data mutually reinforce the recent performance announcements by Japanese MLCC manufacturers, with each company confirming sustained strong order trends. Regarding earnings elasticity, Daiki Takayama estimates that a 5% increase in product prices could theoretically boost Murata Manufacturing's operating profit in FY2027 by approximately 13%, and Taiyo Yuden's operating profit by about 37%. Goldman Sachs forecasts Murata Manufacturing's FY2027 sales at JPY 1.055 trillion (+13% year-on-year) and Taiyo Yuden's at JPY 286 billion (+13% year-on-year). Goldman Sachs has raised its forecast for the year-on-year MLCC price change in 2026 from previously ~0% to a range of 0% to +5%, noting that actual future increases could be far greater. **Another important variable comes from NVIDIA's next-generation Vera Rubin server racks.** According to Morgan Stanley's calculations, the MLCC content value in the VR200 rack is approximately $4,300, a significant increase from the ~$1,500 in the previous generation GB300. Morgan Stanley's channel checks also show that MLCC usage on both compute boards and switch boards is rising significantly, with the increase on compute boards being more pronounced. The newly introduced BlueField and ConnectX modules will further raise the total MLCC content per rack. This partly explains why current demand for high-end AI server MLCCs is so robust, prompting ODM manufacturers to actively build up inventory in preparation for mass production and delivery of Rubin racks starting in the second half of 2026. ### Related Stocks - [6971.JP](https://longbridge.com/en/quote/6971.JP.md) - [6762.JP](https://longbridge.com/en/quote/6762.JP.md) - [6976.JP](https://longbridge.com/en/quote/6976.JP.md) - [TTDKY.US](https://longbridge.com/en/quote/TTDKY.US.md) - [MRAAY.US](https://longbridge.com/en/quote/MRAAY.US.md) - [KYOCY.US](https://longbridge.com/en/quote/KYOCY.US.md) - [6981.JP](https://longbridge.com/en/quote/6981.JP.md) - [GS.US](https://longbridge.com/en/quote/GS.US.md) - [6676.JP](https://longbridge.com/en/quote/6676.JP.md) - [NVDA.US](https://longbridge.com/en/quote/NVDA.US.md) - [MS.US](https://longbridge.com/en/quote/MS.US.md) - [W4VR.SG](https://longbridge.com/en/quote/W4VR.SG.md) - [PJX.SG](https://longbridge.com/en/quote/PJX.SG.md) - [NVD.DE](https://longbridge.com/en/quote/NVD.DE.md) - [MS-O.US](https://longbridge.com/en/quote/MS-O.US.md) - [MS-Q.US](https://longbridge.com/en/quote/MS-Q.US.md) - [MS-E.US](https://longbridge.com/en/quote/MS-E.US.md) - [MS-I.US](https://longbridge.com/en/quote/MS-I.US.md) - [MS-L.US](https://longbridge.com/en/quote/MS-L.US.md) - [MS-P.US](https://longbridge.com/en/quote/MS-P.US.md) - [MS-A.US](https://longbridge.com/en/quote/MS-A.US.md) - [MS-F.US](https://longbridge.com/en/quote/MS-F.US.md) - [MS-K.US](https://longbridge.com/en/quote/MS-K.US.md) ## Related News & Research - [TDK to acquire Malaysia’s lithium-ion rechargeable batteries firm Linergy Power at $240M](https://longbridge.com/en/news/287029149.md) - [A Peek at Kyocera's Future Earnings](https://longbridge.com/en/news/274321905.md) - [Bank of America Securities Keeps Their Sell Rating on Taiyo Yuden Co., Ltd. (6976)](https://longbridge.com/en/news/287097659.md) - [Goldman Sachs Keeps Their Buy Rating on Kyocera (KYOCF)](https://longbridge.com/en/news/274984837.md) - [Murata Manufacturing Co (MRAAF): New Buy Recommendation for This Technology Giant](https://longbridge.com/en/news/284207794.md)