[Edaily Reporter Shin Yeong-bin ] NVIDIA is significantly expanding its global supply chain production capacity to meet the surging demand for artificial intelligence (AI) infrastructure.

NVIDIA announced on the 28th that it is expanding key manufacturing capabilities—including printed circuit boards, advanced packaging, semiconductor testing, and optical engines—in collaboration with its supply chain partners.

First, over the next three years, the company will increase production capacity for printed circuit boards (PCBs) used in NVIDIA products by 2.5 times the current level. Japan’s Ibiden, SamsungElectroMechanics(009150), and Taiwan’s Unimicron are participating in this expansion.

SamsungElectroMechanics will newly join NVIDIA’s substrate supply chain by utilizing the available production capacity at its existing domestic facilities. The company plans to upgrade its production facilities in Sejong and Busan and is also moving forward with the construction of a new factory.

According to NVIDIA, the volume allocated to NVIDIA through 2028 is expected to account for a significant portion of SamsungElectroMechanics’ total annual production capacity for package substrates.

NVIDIA is providing Ibiden with billions of dollars in advance payments to secure production capacity to support the expansion of its Ono plant in Japan. The plan is to expand the plant’s production capacity to three times its current level and gradually secure additional volumes starting in the first quarter of 2028.

Unimicron is also expanding its substrate production capacity in Taiwan. The additional capacity is expected to be secured starting in the second quarter of 2027 and reach its peak by mid-2028.

In the advanced packaging sector, NVIDIA is expanding production capacity in South Korea in partnership with Amkor. The two companies aim to increase interposer wafer production capacity by more than 300% by the end of 2028. This effort, in conjunction with Amkor’s construction of a new manufacturing facility in South Korea, will support the expansion of production for NVIDIA’s current and next-generation AI platforms.

Semiconductor testing capacity will also be expanded. NVIDIA projects that its total testing capacity will double between 2026 and 2027 through capacity expansions in the U.S., Taiwan, and Singapore.

NVIDIA will provide a $1.1 billion advance payment to Taiwan-based King Yuan Electronics (KYEC) to secure production capacity. Based on this, the company aims to build new testing facilities and equipment in the U.S. and begin mass production in April 2027.

Production of optical engines for Co-Packaged Optics (CPO) will also be expanded. NVIDIA and Siliconware Precision Industries (SPIL) plan to increase production capacity in Taiwan 15-fold by mid-2027 and then double it again by the end of 2028.

Jensen Huang, CEO of NVIDIA, said, “We are in the midst of the largest infrastructure build-out in human history,” adding, “Demand is enormous, and production capacity is currently the biggest constraint.”