Samsung Electro-Mechanics and Nvidia are stepping up cooperation on component supplies to respond to a sharp rise in global demand for AI infrastructure. Samsung Electro-Mechanics said on Tuesday it will supply advanced package substrates for computing and high-speed networking products to support Nvidia’s expanded production of its accelerated computing platforms. The companies announced a strategic partnership in October to supply semiconductor package substrates and have now broadened the scope of cooperation to include multilayer ceramic capacitors, or MLCC.
Samsung Electro-Mechanics will continue to expand supplies of advanced package substrates used in Nvidia’s next-generation GPUs. It will also strengthen operational capabilities at its existing plants in Sejong and Busan to raise operating efficiency, maintain a quality management system and respond flexibly to demand fluctuations. Based on packaging process technologies such as implementing ultra-fine circuits, maintaining flatness and embedding passive components, it will supply substrates for high performance computing and for data center networks.
Nvidia agreed in October, when it announced the partnership, to provide large-scale funding to expand dedicated plants in Sejong and Busan for AI semiconductor package substrates. It said it is now reviewing additional funding to expand production infrastructure.
The additional funding will be used to introduce state-of-the-art production equipment, and Samsung Electro-Mechanics plans to establish a full-fledged mass production system from 2029. Nvidia will preemptively secure an advanced substrate supply chain to resolve supply bottlenecks, while Samsung Electro-Mechanics will expand its high value-added package substrate business.
The scope of cooperation will also expand from substrates to MLCC. AI accelerated computing platforms require stable power supply as a key condition in highly integrated, high-output computing environments. Samsung Electro-Mechanics will also move forward with cooperation on MLCC needed for Nvidia’s computing platforms, based on its high-reliability MLCC technology.
Debora Shoquist (데보라 쇼퀴스트), Nvidia’s executive vice president of operations, said, "To meet long-term demand for AI infrastructure, close cooperation across the global supply chain is essential." She added, "Our collaboration with Samsung Electro-Mechanics builds on our existing substrate supply partnership and is intended to support expanded production of Nvidia’s current and next-generation accelerated computing platforms."
Chang Duck-hyun (장덕현), CEO of Samsung Electro-Mechanics, said, "By combining world-class substrate fine-processing technology with high-reliability MLCC capabilities, we will flawlessly fulfill our role as a total solution provider so Nvidia’s AI platforms can be supplied to the global market in a timely manner with the best performance and quality."