What the agreement entails

Samsung Electro-Mechanics and Nvidia are broadening their collaboration in AI components. The Korean company announced this on Wednesday, according to The Korea Herald: the agreement, until now focused on substrates for AI chip packaging, is also being extended to multilayer ceramic capacitors, known as MLCCs. These are tiny electronic components that help stabilize the power supply to processors and other circuits.

The basis for the agreement dates back to October, when Samsung Electro-Mechanics agreed to supply Nvidia with advanced substrates for AI processors and for high-speed data center networking chips. The expansion announced now builds on that strategic partnership and extends its scope to a new family of components.

Production capacity and financial support

According to the Korean newspaper's account, Nvidia is considering additional funding to increase production capacity at Samsung Electro-Mechanics' plants in Sejong and Busan and to install advanced manufacturing equipment at the two sites. Already at the time of the October announcement, the American manufacturer had committed substantial financial support to expand the lines dedicated to AI chip substrates. The stated rationale is the need to secure more components for its growing AI-based computing business.

Under the expanded agreement, Samsung Electro-Mechanics will increase supplies of advanced substrates for Nvidia's next-generation GPUs, leveraging its manufacturing operations in Sejong and Busan. The company plans to improve production efficiency, strengthen quality controls and make the lines more flexible in response to demand fluctuations. It also intends to leverage advanced packaging technologies such as ultra-fine circuit processing, substrate planarity control and integrated passive components to supply high-performance computing and high-speed networking systems. The company aims to establish full-scale mass production capacity starting in 2029, supported by the expanded facilities.

The role of MLCCs and the statements

The extension to MLCCs is presented as a significant step because these components are gaining increasing importance in AI computing platforms, where high-performance, high-density chips require reliable power supply. Samsung Electro-Mechanics intends to use its high-reliability capacitor technology to continue cooperation with Nvidia and expand its role in the American company's AI hardware supply chain.

Samsung Electro-Mechanics CEO Chang Duck-hyun said that by combining fine-processing technology for substrates and high-reliability MLCC technology, the company intends to fully play the role of a complete solutions provider, so that Nvidia's AI platform can reach the global market on schedule, with maximum performance and quality.

During a Tuesday earnings conference call, Colette Kress, Nvidia's executive vice president and chief financial officer as well as interim CEO, said the company had signed new substrate agreements with Samsung Electro-Mechanics and other suppliers. In a separate communication, Nvidia announced that it had amended its existing agreement with the Korean company to improve production capacities and support increased output of its accelerated computing platforms.

What remains to be verified

The material does not specify the amount of the additional funding under consideration or the timing of any formalization, and does not indicate the overall value of the agreement. The impact on supply volumes is also not quantified. The information comes from company announcements and statements reported in the press, without independent confirmation of the transaction's economic details.

Source and attribution: this original summary was prepared by AICronaca based on the report by The Korea Herald ↗. The statements are attributed to the source; possible information limitations: The article is based solely on material provided by The Korea Herald and on the statements reported in it. No amounts for the additional funding, agreement values, supply volumes or independent confirmations are available..