Test report DSG-7443 · Rev D · tested November 18, 2026
Supply Chain & PolicyDevice under test
Samsung Electro-Mechanics Expands Substrate Output With Nvidia Funds
Samsung Electro-Mechanics will expand package substrate capacity for Nvidia's AI chips, with full-scale output at Sejong and Busan from 2029 and possible extra Nvidia funding.
- Read
- 3 min
- Words
- 502
- Node
- 28nm
- Operator
- Grace Kim
Spec summary
- Samsung Electro-Mechanics targets full-scale package substrate output at Sejong and Busan in 2029
- Nvidia announced the contract revision on its Q3 earnings call on Nov. 17
- Nvidia is reviewing additional funding on top of last month's large-scale financing
- Cooperation extends to high-reliability MLCCs for Nvidia AI platforms

Samsung Electro-Mechanics (KRX: 009150) will expand package substrate capacity for Nvidia's next-generation AI chips, with full-scale output at its Sejong and Busan plants targeted for 2029. Nvidia is reviewing a plan to provide additional funding beyond the large-scale financing it already committed last month.
The Korean component maker announced the move on Nov. 18. It plans to upgrade operating systems at both plants in line with Nvidia's expanded production of AI accelerated computing platforms. Nvidia confirmed its side of the deal one day earlier, saying on its third-quarter earnings call on Nov. 17 local time that it would revise its existing package substrate supply contract with Samsung Electro-Mechanics to expand capacity.
Package substrates connect a chip to its motherboard and carry its power and signal lines. They're a bottleneck component in AI hardware, which is why Nvidia is paying to secure supply.
What does the expansion cover?
The plan has three main elements, according to the company:
- Enlarged production facilities at Sejong and Busan
- Installation of state-of-the-art equipment, funded in part by Nvidia
- Upgraded operating systems allowing output to flex up or down with demand shifts
Samsung Electro-Mechanics said it will also raise the operating efficiency of its existing package substrate lines, not just add new capacity.
Deeper Nvidia money, wider cooperation
Last month the two companies signed a strategic partnership for substrate supply. Under that deal, Nvidia agreed to provide large-scale funding for dedicated AI chip substrate plants in Sejong and Busan. The additional funds now under review would flow into the same two sites, paying for bigger facilities and advanced equipment.
The cooperation is also widening to multilayer ceramic capacitors, or MLCCs. These components stabilize the voltage supplied to chips when power consumption swings sharply with computing loads — a daily reality for AI accelerators. Samsung Electro-Mechanics said it will work with Nvidia on high-reliability MLCCs required for the US company's AI computing platforms.
"We will combine the world's best substrate micro-processing technology with high-reliability MLCC capabilities," said Chang Duck-hyun, president of Samsung Electro-Mechanics. "We will fully carry out our role as a total solutions provider so that Nvidia's AI platforms can be supplied to the global market on time with the highest performance and quality."
Nvidia framed the deal as continuity, not a fresh start. Debora Shoquist, the company's executive vice president of operations, said the expanded cooperation builds on the existing package substrate partnership and is intended to support the scaling of production for current and next-generation accelerated computing platforms.
What does this change for the supply chain?
For Samsung Electro-Mechanics, it locks in a committed buyer and external funding for a capacity ramp through 2029. For Nvidia, it secures a second pillar of component supply — substrates plus MLCCs — from a single Korean partner as it scales AI platform output. No investment amounts, capacity figures or hiring numbers were disclosed by either company.
via en.sedaily.com (Source)
More from Grace Kim
Show full bio
Market editor covering regulation, policy and their impact on hardware layer of artificial intelligence at Die Signal.
665 articles
Same lot · LOT-362E
- DSG-15343nmNVIDIA Targets 2.5x Substrate Capacity, Doubles Testing by 2027
- DSG-754020nmNvidia Signs New Substrate Deals With Samsung Electro-Mechanics, Kinsus
- DSG-820310nmNvidia reserves AI accelerator capacity through 2028
- DSG-42087nmHanwha sees GPU shortage lasting through Nvidia’s fiscal 2028
- DSG-26535nmSamsung Electro-Mechanics Wins 360.5 Billion Won Chip Substrate Order