Additional Funding for Sejong and Busan Sites Under Consideration
Full-Scale Production of Substrates for Next-Generation GPUs to Begin in 2029

Samsung Electro-Mechanics and NVIDIA are stepping up their strategic cooperation to meet surging global demand for artificial intelligence (AI) infrastructure.


The two companies announced plans to expand their cooperation on November 18 to support increased production of NVIDIA’s accelerated computing platforms. Last month, they signed a strategic partnership to supply semiconductor package substrates. Under the expanded cooperation, Samsung Electro-Mechanics will provide a stable supply of advanced package substrates for NVIDIA’s next-generation graphics processing units (GPUs) and high-speed networking products.

Samsung Electro-Mechanics, NVIDIA Expand AI Infrastructure Cooperation... Alliance Extends from Substrates to MLCCs View original image

To this end, Samsung Electro-Mechanics will enhance the operational capabilities of its existing plants at its key domestic manufacturing sites in Sejong and Busan. It will improve production-line efficiency while maintaining a quality control system at the highest level. The company will also establish a flexible production system that can respond immediately to changes in demand. Samsung Electro-Mechanics has cutting-edge packaging process technologies that enable ultra-fine circuit patterns, maintain flatness, and embed passive components. Building on these capabilities, it plans to supply package substrates for networking products that support high-performance computing (HPC) and ultra-high-speed data transmission in data centers.


The companies are also expanding their financial cooperation to build out production infrastructure. When they announced their partnership last month, NVIDIA agreed to provide substantial funding for the expansion of dedicated AI semiconductor package substrate plants at the Sejong and Busan sites. To support the expanded cooperation, NVIDIA is also considering additional funding to expand production infrastructure and introduce state-of-the-art manufacturing equipment at the two sites. With this funding, Samsung Electro-Mechanics plans to establish a full-scale mass production system starting in 2029.


Through this cooperation, NVIDIA will secure in advance a supply chain for state-of-the-art substrates essential to mass-producing AI semiconductors, helping ease bottlenecks in the global AI infrastructure supply chain. Samsung Electro-Mechanics will substantially expand its business in high-value-added AI semiconductor package substrates, a field with high barriers to entry, and strengthen its position in the global market.


The scope of the partnership will expand beyond package substrates to multilayer ceramic capacitors (MLCCs). A stable power supply is essential for AI accelerated computing platforms that perform high-density, high-power computations. Drawing on its world-class, highly reliable MLCC technology, Samsung Electro-Mechanics will also work with NVIDIA to supply the MLCCs required for its computing platforms. The company is expected to consolidate its position as a total solutions provider for AI components, offering both substrates and MLCCs.


MLCCs regulate the flow of electricity in electronic product circuits to ensure it remains steady and stable. They are made by stacking hundreds of layers of dielectric material and electrodes inside an ultra-small chip.


“By combining world-class fine substrate processing technology with highly reliable MLCC technology, we will fully fulfill our role as a total solutions provider, ensuring that NVIDIA’s AI platforms reach global markets on time with the highest performance and quality,” said Jang Deok-hyun, president of Samsung Electro-Mechanics.



“Close cooperation across the global supply chain is essential to meeting the long-term demand for AI infrastructure,” said Deborah Shoquist, executive vice president of operations at NVIDIA. “Our cooperation with Samsung Electro-Mechanics builds on our existing substrate supply partnership and is intended to support increased production of NVIDIA’s current and next-generation accelerated computing platforms.”


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing