
Samsung Electro-Mechanics said on Nov. 18 that it will increase supplies of advanced packaging substrates for Nvidia's next-generation graphics processing units (GPUs) and expand its partnership to include multilayer ceramic capacitors (MLCCs), supporting the ramp-up of Nvidia's accelerated computing platforms.
The move builds on a packaging substrate supply partnership signed last month. Samsung Electro-Mechanics will leverage advanced packaging technologies, including ultrafine circuitry, to supply substrates for Nvidia's high-performance computing (HPC) and data center networking applications. To support the effort, the company will improve production-line efficiency and strengthen quality control at its Sejong and Busan facilities.
Further investment in production infrastructure is also planned. Having committed funding for the expansion of a dedicated factory, Nvidia is considering additional financial support for equipment at the Sejong and Busan sites. Samsung Electro-Mechanics plans to begin full-scale mass production in 2029. The companies expect the partnership to help Nvidia ease substrate supply bottlenecks for AI chips while strengthening Samsung Electro-Mechanics' position in the high-value substrate market, where barriers to entry are high.
The partnership will also extend to high-reliability MLCCs—compact electronic components that help stabilize power delivery. A stable power supply is essential for AI accelerated computing platforms operating in dense, power-intensive environments. Samsung Electro-Mechanics plans to build on its expertise in high-reliability MLCCs to collaborate on components for Nvidia's computing platforms, positioning itself as a comprehensive supplier of AI hardware components.
“By combining our fine-pattern substrate manufacturing technology with our expertise in high-reliability MLCCs, we will provide comprehensive solutions that help ensure the timely supply of Nvidia's AI platforms worldwide,” said Chang Duck-hyun, president and CEO of Samsung Electro-Mechanics.
