NVIDIA, Samsung Electro-Mechanics and other partners step up cooperation... expanding the AI semiconductor supply chain

[Financial News] NVIDIA is expanding its global production capacity to meet growing demand for AI computing. It plans to broaden its production base and strengthen supply chain resilience by deepening cooperation with supply chain partners in key areas, including substrates, advanced packaging, semiconductor testing and optical engines.
NVIDIA said on the 28th that it would work with Amkor Technology, Ibiden, King Yuan Electronics (KYEC), Samsung Electro-Mechanics, Siliconware Precision Industries (SPIL) and Unimicron to actively support capacity expansions at both new and existing suppliers.
NVIDIA founder and CEO Jensen Huang said, "The largest infrastructure buildout in human history is underway, and computing demand is surging, but the biggest constraint today is production capacity. We are grateful to our partners for working with us to scale production around the world and deliver AI supercomputing systems."
NVIDIA plans to increase substrate production capacity for its products to 2.5 times its current level over the next three years. Under an agreement announced early this year, Ibiden will expand capacity at its Ono plant in Japan to about three times its current level. NVIDIA is providing several billion dollars in capacity prepayments to support the expansion. Additional capacity is expected to come online in stages starting in the first quarter of 2028.
Samsung Electro-Mechanics will newly join NVIDIA's substrate supply chain, using available capacity at its existing production facilities in Korea. The collaboration includes upgrades to production facilities in Sejong and Busan, as well as the construction of a new plant. By 2028, production allocated to NVIDIA is expected to account for a significant share of Samsung Electro-Mechanics' total annual package substrate capacity.
Unimicron will expand its KF substrate production capacity in Taiwan. NVIDIA is supporting the expansion through a capacity reservation agreement. Additional capacity is expected to start becoming available in the second quarter of 2027 and reach its maximum level by mid-2028.
NVIDIA and Amkor are expanding advanced packaging capacity in Korea. The two companies aim to increase interposer wafer capacity by at least 300% by the end of 2028. The expansion is expected to support increased production of NVIDIA's current and next-generation platforms, alongside the construction of Amkor's new manufacturing facility in Korea.
NVIDIA expects its overall semiconductor testing capacity to double between 2026 and 2027, driven by capacity expansions in the U.S., Taiwan and Singapore.
The $1.1 billion capacity reservation prepayment to KYEC will fund the construction of a new testing facility and the installation of equipment in the U.S. Mass production at the facility is targeted to begin in April 2027. In Singapore, the expanded testing facility is scheduled to begin mass production in February 2027.
NVIDIA and SPIL are increasing production capacity for co-packaged optics (CPO) optical engines in Taiwan to meet the connectivity needs of large-scale computing infrastructure. The goal is to expand capacity to 15 times its current level by mid-2027 and double it again by the end of 2028. 
[email protected] Yoon Hong-jip Reporter