Economy/Home · Economy

Samsung Electro-Mechanics Introduces MLCC Technology for 1MW-Per-Rack AI Server Power Challenges... GPU Power Stabilization Emerges as Key Task

As power consumption in artificial intelligence (AI) servers rapidly increases, the importance of electronic components responsible for stable power supply—alongside semiconductors—is growing. In particular, in server environments where high-performance graphics processing units (GPUs) are densely i

Wooil Shim
Staff Reporter
10 min read
Samsung Electro-Mechanics Introduces MLCC Technology for 1MW-Per-Rack AI Server Power Challenges... GPU Power Stabilization Emerges as Key Task
CBC News

As power consumption in artificial intelligence (AI) servers rapidly increases, the importance of electronic components responsible for stable power supply—alongside semiconductors—is growing. In particular, in server environments where high-performance graphics processing units (GPUs) are densely integrated, supplying large amounts of power in a limited space has brought voltage stability and component miniaturization to the forefront as core technical challenges.

Samsung Electro-Mechanics announced on the 16th, through its official social media, multilayer ceramic capacitor (MLCC) technology designed for the era of 1MW (megawatt)-class AI server racks. 1MW is equivalent to 1,000kW, a figure that illustrates the trend of increasing power density in AI server racks.

◆ What Is an MLCC... A Key Passive Component for Voltage Stabilization and Noise Reduction

The change Samsung Electro-Mechanics highlighted is the restructuring of power supply systems driven by improved GPU performance. As the number of GPUs processing AI computations grows and their power consumption increases, the company explained that technologies for controlling voltage fluctuations and noise generated during power conversion are becoming more critical.

MLCCs temporarily store and release electricity in electronic circuits, maintaining stable voltage and reducing noise. Though small in size, they are one of the key passive components necessary for the stable operation of high-performance semiconductors and power supply circuits.

◆ Three Key Technology Directions Presented by Samsung Electro-Mechanics

Samsung Electro-Mechanics presented three main directions for AI server MLCC technology: ▲support for high-voltage DC-DC conversion circuits, ▲optimization of GPU core power, and ▲embedded MLCCs in circuit boards.

First, in high-voltage DC-DC conversion circuits, instantaneous voltage fluctuations and spike noise can occur during rapid power switching. Samsung Electro-Mechanics explained that MLCCs can help reduce such electrical instability.

The role of MLCCs in GPU core power supply is also expanding. Samsung Electro-Mechanics noted that the MLCC capacitance required per GPU is doubling every year. However, this reflects a technology demand trend presented by the company and is difficult to generalize as a figure that applies equally to all GPU products.

In power supply circuits, appropriately combining MLCCs with different characteristics is more important than simply using components with large capacitance, because the required capacitance and electrical characteristics vary depending on the GPU's power consumption profile and circuit design.

Board-embedded MLCCs are also a technology worth watching. Placing MLCCs inside the circuit board, rather than on the surface as before, allows efficient use of limited space and can also help optimize component placement and power delivery paths.

◆ Power Efficiency and System Stability Determine Competitiveness

These technologies are gaining importance in line with the trend toward higher integration in AI servers. As GPU performance improves, circuit design for stable power supply becomes more complex, requiring a high level of technical capability from related components.

In particular, in the AI data center industry, not only computing performance but also power efficiency and system stability are emerging as factors that determine competitiveness. Even with sufficient high-performance semiconductors, unstable power supply makes it difficult to maintain system performance reliably. This shift in the industrial environment underlies Samsung Electro-Mechanics' emphasis on MLCC technology.

As power density in AI servers increases, the importance of electronic components with high-voltage handling capabilities, miniaturization technology, and power stabilization performance is likely to grow.

◆ Link to Earnings Remains Uncertain... Adoption and Supply Volume Are Key

However, it cannot be assumed that growth in the AI server market will directly translate into increased MLCC revenue for Samsung Electro-Mechanics. The actual impact on earnings may vary depending on customer adoption, supply volumes, product mix, pricing, and profitability.

Going forward, key factors to watch in Samsung Electro-Mechanics' AI server MLCC business will include whether supply of high-value-added products expands, customer acquisition, and the commercial results of its technological competitiveness. As competition in the AI industry expands beyond semiconductor computing power to power supply and component integration technology, attention is focused on what role Samsung Electro-Mechanics' MLCC technology will play.

[This article was written based on Samsung Electro-Mechanics' official technology introduction and does not guarantee the performance or stock price of any particular company. Industry outlooks and actual business results may differ, and investment decisions and responsibilities rest with the investor. AI assistance was used in writing this article.]

Wooil Shim
Staff Reporter

CBC Globe publishes verified stories with editorial review, source checks, and tenant-specific publication standards.