Thursday, October 22, 2026

Samsung Electro-Mechanics becomes NVIDIA’s “AI chip buddy,” builds dedicated production line in South Korea

Input
2026-10-21 20:30:00
Updated
2026-10-21 20:30:00
NVIDIA CEO Jensen Huang speaks at the opening ceremony of the Germany Investment Summit (Invest in Germany Summit) in Berlin, Germany, on the 19th (local time). “Digital technology will not replace traditional manufacturing; it will be a means of enhancing its competitiveness,” Huang said. “Just as electrons cannot replace atoms, artificial intelligence (AI) will only enhance our capabilities, not replace them. Therefore, we should make active use of this technology.” Berlin = AP/Newsis

A view of Samsung Electro-Mechanics’ Suwon campus. Provided by Samsung Electro-Mechanics.

[Financial News] Samsung Electro-Mechanics is joining forces with NVIDIA, a leader in the global AI semiconductor market, to build a production line in South Korea dedicated to AI semiconductor package substrates. The two companies will jointly invest to expand production capacity at their Sejong and Busan sites and establish a multi-year, long-term supply arrangement covering substrates for both AI accelerators and data-center networks. The move is expected to strengthen Samsung Electro-Mechanics’ position as a key supplier in the global AI semiconductor supply chain, alongside the record 4.27 trillion won investment in substrate capacity expansion it announced last September.■ Building a production line dedicated to NVIDIAAccording to industry sources on the 21st, Samsung Electro-Mechanics has formed a strategic business partnership with NVIDIA to strengthen the global AI semiconductor supply chain and accelerate the expansion of next-generation AI infrastructure. Under the agreement, the two companies will jointly invest to expand facilities for producing AI semiconductor package substrates. Samsung Electro-Mechanics plans to supply NVIDIA with package substrates for AI semiconductors on a long-term basis over multiple years.
NVIDIA will secure a stable supply chain in advance for the cutting-edge substrates essential to mass-producing AI semiconductors, helping ease bottlenecks in the supply of global AI infrastructure. Samsung Electro-Mechanics, meanwhile, will step up its high-value-added AI semiconductor package substrate business, where technological barriers to entry are the highest, and strengthen its position in the global market.
Specifically, Samsung Electro-Mechanics will establish new dedicated production lines at its Sejong and Busan sites, key domestic manufacturing hubs, to ensure timely delivery of semiconductor package substrates. Products to be supplied include substrates for NVIDIA’s computing chipsets and package substrates for network (switch) equipment, which enables ultrafast connections between chips in data centers. Supply is set to begin this year. Starting this year, the volume supplied to NVIDIA by Samsung Electro-Mechanics’ package substrate business is expected to grow by double digits or more on an annual average basis over multiple years. By 2028, NVIDIA’s orders are expected to account for a substantial share of Samsung Electro-Mechanics’ total annual package substrate production capacity.
Sales at Samsung Electro-Mechanics’ Package Solutions Division, which produces semiconductor substrates for AI servers and other products, reached 771.6 billion won in the second quarter of this year, up 37% from the same period a year earlier. The division continues to grow on the back of rising demand for high-value-added package substrates for AI servers. Sales growth is expected to gain further momentum as volumes under the long-term supply agreement with NVIDIA ramp up.■ Expanding the supply base with a 4.3 trillion won investmentThe partnership with NVIDIA also comes alongside Samsung Electro-Mechanics’ major investment, announced last September, to expand its package substrate production facilities. On September 28, Samsung Electro-Mechanics said it would invest a total of 4.27 trillion won at its Sejong site to expand production capacity for substrates including flip-chip ball grid array (FC-BGA) products for AI servers. It is the company’s largest-ever investment for a single product. The company plans to complete the expansion by May 2028, start operating the new production line in September of that year, and begin mass production. Samsung Electro-Mechanics previously said the investment was being pursued on the basis of mid- to long-term cooperation with global big-tech customers. It also explained that the arrangement included customer funding support and guarantees of supply volumes over the medium to long term. The announcement of the strategic partnership with NVIDIA is seen as offering a clearer picture of the global customer collaboration linked to the large-scale capital investment.
Samsung Electro-Mechanics CEO Jang Deok-hyun said, “We are honored to join as a key infrastructure partner to NVIDIA, which is leading global AI innovation. By combining our accumulated FC-BGA manufacturing expertise with our unrivaled, highly reliable multilayer ceramic capacitor (MLCC) technology, we will fully fulfill our role as a total solutions provider, ensuring NVIDIA’s AI platforms reach the market on time with the highest performance and quality.”
NVIDIA Executive Vice President of Operations Deborah Shoquist said, “To keep pace with the speed of AI infrastructure buildout, production capacity must be expanded globally. Our collaboration with Samsung Electro-Mechanics has broadened our network of substrate supply partners and secured additional production capacity. This will allow us to scale up production and supply the accelerated computing platforms our customers need.”
[email protected] Lim Su-bin Reporter