Samsung Electro-Mechanics set to gain from AI demand

KB Securities said on May 4 that Samsung Electro-Mechanics could stronger results as AI demand lifts sales of multilayer key components.

The brokerage maintained a buy rating and raised its target price, citing demand for multilayer ceramic capacitors and FC-BGA substrates used in AI servers and power systems.

Samsung Electro-Mechanics reported first-quarter revenue of 3.2 trillion won (US$2.18 billion) and operating profit of 280.6 billion won (US$191 million), up 17% and 40% from a year earlier. KB Securities said the results beat expectations after excluding a one-off retirement cost.

The company expects second-quarter revenue of 3.3 trillion won (US$2.23 billion) and operating profit of 398.3 billion won (US$271 million), above analyst estimates, supported by a higher share of margin-rich AI components.

🔗 Source: The Chosun Daily

🧠 Food for thought

Implications, context, and why it matters.

Samsung Electro-Mechanics’ AI component demand is outstripping supply by up to 100%

Samsung Electro-Mechanics’ recent results rest on a sharp gap between demand and supply for its most advanced parts 1.
Demand for the company’s server flip-chip ball grid array (FC-BGA) substrates, which link AI processors to circuit boards, is more than 50% above production capacity 1.
For high-end multilayer ceramic capacitors (MLCCs), small parts that regulate power flow in devices, customer inquiries at leading suppliers are running at about twice available capacity 2.
An AI server board can use ten to twenty times more MLCCs than a conventional smartphone, turning a once-commoditized part into a bottleneck 3.

AI’s new bottlenecks are the unsung components holding chips together

Murata and Taiyo Yuden, Japanese electronic component makers, are also moving toward price increases, which suggests pricing power is shifting to a small group of suppliers 4.
Limited supply of high-end FC-BGA substrates can affect competition between chipmakers such as AMD and Intel 5.
Unlike earlier component swings tied to uneven smartphone orders, this rise comes from AI data center buyers with two-to-three-year purchasing roadmaps 4.
That demand pattern could keep pricing firm for component makers and shape the wider AI hardware supply chain for years 2.