Samsung Electro-Mechanics projected that supply and demand for multilayer ceramic capacitors (MLCCs) will tighten further as investment in artificial intelligence (AI) infrastructure expands. In response, it plans to expand long-term supply agreements (LTAs) with major big tech corporations while continuing price negotiations.
Samsung Electro-Mechanics said during its second-quarter earnings conference call on the 30th that "demand for MLCCs is steadily increasing across various applications, including AI and the automotive electronics industry," adding, "we are actively pursuing discussions on long-term supply agreements with major big tech customers."
The company also predicted that supply and demand for titanium (Ti), a core MLCC materials and supplies, will become even tighter as AI demand grows.
Samsung Electro-Mechanics explained, "as the titanium supply-demand situation is expected to intensify due to rising AI demand, we plan to closely monitor conditions and continue a strategic pricing response."
By application, it projected that demand for MLCCs for AI servers and data centers will continue to rise. In the automotive electronics institutional sector, demand is increasing for compact, ultra-high-capacitance MLCCs as advanced driver-assistance systems (ADAS) become more sophisticated, and it also forecast growth in the market for 100V-class high-capacitance MLCCs for 48V power systems and high-voltage products for ultra-fast charging infrastructure.
IT MLCCs are also expected to benefit from the spread of AI Semiconductor and next-generation memory. Samsung Electro-Mechanics said it will expand supplies of ultra-compact, ultra-high-capacitance MLCCs and low-ESD products for mobile application processor (AP) packages to major semiconductor companies, and will actively target the market for high-temperature, low-profile MLCCs used in next-generation memory packages for DDR6 and LPDDR6.
It also projected that the semiconductor substrate (FC-BGA) market will continue to face supply shortages.
Samsung Electro-Mechanics assessed that as the market expands from Generative AI to agentic AI, demand for server central processing units (CPUs) and AI accelerators is maintaining solid growth. Accordingly, it has begun full-scale supply of high-value substrates for AI data center networks to major North American big tech corporations, and said data center-related revenue increased sharply from the previous quarter.
The company said, "as data center semiconductors become larger in area and have more layers, FC-BGA supply shortages are expected to deepen further," adding, "we will expand supplies of high-value substrates for AI data centers based on multicore and embedding technologies, and secure early market leadership through capacity expansion and long-term supply agreements."
In particular, it emphasized, "in the third quarter as well, demand for server CPUs and AI accelerators is expected to remain solid, keeping the market in favor of suppliers," adding, "we will continue price talks with major customers while expanding advanced substrate capacity through new plant expansions."