An AI server consumes 15 times more capacitors: MLCC, substrates, and TCB may become new winners in the chip supply chain.

```

The wave of AI infrastructure construction is spreading deep into the supply chain, accelerating the entry of previously little-known component manufacturers into investors’ sights.

As computing power demand continues to surge, AI servers consume far more electronic components than traditional servers, with the usage gap for some parts reaching dozens of times. According to Young Jae Lee, Senior Investment Manager at Pictet Asset Management, a single AI server consumes 10 to 15 times more multilayer ceramic capacitors (MLCCs) than a standard server, and about 30 times more than a smartphone.

The sharp increase in demand is driving up prices for related products. This week, Samsung Electro-Mechanics said it is considering increasing MLCC product prices by up to 10%. Goldman Sachs believes that MLCC demand will grow 4.3 times between 2025 and 2030, and estimates price gains of 0-5%.

The market has already responded. In the past 12 months, the share prices of substrate manufacturers Unimicron and Ibiden have surged about 770% and 530% respectively. MLCC leaders Samsung Electro-Mechanics and Murata Manufacturing both hit historical highs this month; Hanmi Semiconductor, the leader in the field of thermal compression bonding (TCB), also reached a record high.

Three Key Components: From Capacitors to Substrates to Bonding Processes

In this round of AI supply chain boom, investors have focused their attention on three core categories of components.

Multilayer ceramic capacitors (MLCCs) regulate power within electronic systems and are the fundamental components for stable power supply. Advanced chip substrates serve as the medium connecting semiconductors with other hardware. Kieron Poon, Head of Asian Equities at Aberdeen Investments, used a dining table analogy: "Imagine a printed circuit board as a table, the plates on the table are substrates, and the food on the plates are chips." Thermal compression bonding (TCB) is the packaging process that precisely integrates these components, and is crucial for advanced packaging.

The logic driving the surge in demand for these components lies in the high energy consumption of AI infrastructure construction. AI servers consume much more power than traditional servers, and higher power requires more components to manage and stabilize power supply, creating a cascading effect that propels the entire supply chain.

Capacity Utilization Above 90%, Price Pressure Mounts

Constraints on the supply side are beginning to appear. Simon Woo, Head of Korea Research at BofA Global Research, said production line utilization for both MLCCs and substrates has exceeded 90%. "If AI demand rises even slightly, the available capacity for traditional uses will rapidly shrink," he said.

Expectations of price increases are also mounting. Citi analyst Takayuki Naito pointed out that the market is raising expectations for price increases in MLCCs, aluminum electrolytic capacitors, packaging substrates, and other components, believing this will benefit Japanese manufacturers such as Murata Manufacturing and Taiyo Yuden, who may adopt more aggressive pricing. JPMorgan analysts last week raised their price targets for Murata and Taiyo Yuden, and predicted that tight supply-demand conditions will persist for a considerable time.

In terms of competition, supply for these components is highly concentrated, with almost all companies based in Asia, primarily in South Korea, Japan, and China. Meanwhile, the customer base continues to expand alongside AI infrastructure investment. Aberdeen’s Poon believes this means pricing power "will undoubtedly remain in the hands of suppliers."

The ripple effects of this trend are spreading throughout a broader ecosystem. Optical component manufacturers are also caught up in the surge, as demand for bandwidth in data centers driven by AI applications rises, prompting investors to pay closer attention to the optical technology’s role.

Risk Warning and DisclaimerThe market entails risks, and investment should be undertaken with caution. This article does not constitute personal investment advice and does not take into account individual users’ specific investment goals, financial status, or needs. Users should consider whether any opinions, views, or conclusions in this article suit their particular circumstances. Investments made based on this content are at your own risk. ```