Nvidia prepays six suppliers including Samsung Electro-Mechanics and Ibiden to expand production, locking in AI infrastructure capacity
On October 21, Nvidia announced partnerships with Samsung Electro-Mechanics, Ibiden, Amkor, KYEC, SPIL, and Unimicron to lock in packaging substrate, testing, and co-packaged optics (CPO) capacity through prepayments. Among these, the agreements with Ibiden, Samsung Electro-Mechanics, and Unimicron will expand semiconductor packaging substrate capacity 2.5-fold over the next three years; Ibiden's Ono plant in Japan will triple its capacity and reach full production in the first quarter of 2028. Nvidia also prepaid $1.1 billion to KYEC, doubling its total testing capacity in 2027 compared with 2026, and is working with SPIL to increase CPO optical engine capacity 15-fold by mid-2027.
Why read this
AI score 62
The original text lists Nvidia's prepayments to six suppliers and the expansion timelines, which can be used to understand the layout pace of AI server packaging, testing and CPO capacity.
Topics
Body · AI translation
Nvidia is collaborating with Samsung Electro-Mechanics, Ibiden, Amkor Technology, KYEC, SPIL and Unimicron. (Photo: Nvidia)
PC본문_6월 Mouser (in Article)//
AD Manager | AD178089411676 //
// AD Manager | AD178089411676
Nvidia is investing to meet growing demand for artificial intelligence (AI) infrastructure. The company is making advance payments to key suppliers to secure capacity for semiconductor packaging substrates, testing and co-packaged optics (CPO).
Nvidia said Oct. 21 local time it is working with Samsung Electro-Mechanics, Ibiden, Amkor Technology, King Yuan Electronics Co. (KYEC), Siliconware Precision Industries (SPIL) and Unimicron to meet rising demand for AI factories. “We’re in the middle of the largest infrastructure buildout in human history,” Nvidia chief executive officer Jensen Huang said. “Demand is enormous, but capacity is constrained.”
Nvidia is expanding agreements with substrate manufacturers Ibiden, Samsung Electro-Mechanics and Unimicron to increase semiconductor package substrate production capacity 2.5-fold over the next three years. These substrates provide electrical connections and transmit signals between chips and main circuit boards. Flip-chip ball grid array (FC-BGA) substrates used in AI servers are a representative example.
Nvidia has reserved several billion dollars worth of production capacity at Ibiden and made advance payments. The investment aims to triple production capacity at the company’s Ono plant in Japan. The additional capacity is expected to begin full-scale operations in the first quarter of 2028.
Samsung Electro-Mechanics will also begin supplying next-generation semiconductor package substrates from its South Korean production facilities in 2026. Nvidia will support equipment upgrades at Samsung Electro-Mechanics’ Sejong and Busan plants, as well as construction of a new factory.
Nvidia has also signed a capacity reservation agreement with Unimicron to support expansion of its KF plant in Taiwan. Capacity will begin increasing in the second quarter of 2027 and is expected to reach full utilization by mid-2028.
In advanced packaging, Nvidia is working with Amkor to expand production capacity in Gwangju, South Korea. The company plans to increase interposer wafer production capacity by more than 300% by the end of 2028. Interposers serve as intermediate connections between chips and package substrates.
Nvidia is also expanding semiconductor testing capacity. It has committed $1.1 billion in advance payments to reserve capacity at Taiwanese testing company KYEC. Expansion of facilities in the United States, Taiwan and Singapore will double total semiconductor testing capacity in 2027 from the 2026 level. A new U.S. testing facility is scheduled to begin mass production in April 2027, while expanded production lines in Singapore will begin mass production in February 2027.
Nvidia is also working with SPIL to expand optical engine production capacity for CPO in Taiwan. Capacity will increase 15-fold by mid-2027 and double again by the end of 2028. The expansion aims to secure next-generation optical signal components.