Nvidia Multiplies Agreements to Relieve Production Capacity Constraints
At the GTC Berlin conference on October 21, 2026, the group presented a series of agreements aimed at expanding substrate production, advanced packaging, semiconductor testing, and optical engines. Jensen Huang, founder and chief executive officer, summarized the challenge: "Demand is extraordinary. Capacity is the constraint."
Substrates: Target of 2.5x Capacity Increase Within Three Years
Nvidia announced projects with Amkor Technology, Ibiden, King Yuan Electronics (KYEC), Samsung Electro-Mechanics, Siliconware Precision Industries (SPIL), and Unimicron Technology Corporation. The stated objective is to meet growing demand for its accelerated computing platforms and AI factories, while strengthening "resilience and redundancy" in its global supply chain. The group indicates it is supporting both existing suppliers and new entrants.
On substrates, Nvidia plans to increase its capacity by 2.5x over the next three years, with additional agreements in progress. Under an agreement signed and announced earlier in the year by Ibiden, Nvidia provides a multi-billion-dollar capacity reservation advance for the Ono plant in Japan, whose capacity is expected to reach three times its current level. Additional capacities are scheduled to ramp up in the first quarter of 2028.
Samsung Electro-Mechanics is joining the substrate supply chain, with immediate capacity from its existing sites in South Korea, as well as upgrades and new factories in Sejong and Busan. By 2028, production allocated to Nvidia is expected to represent a significant share of the South Korean group's total annual substrate packaging capacity. In Taiwan, a capacity reservation agreement supports the expansion of Unimicron's KF substrate business, with initial capacities expected in the second quarter of 2027 and full capacity targeted for mid-2028.
Packaging, Testing, and Co-integrated Optics: Timelines Through 2028
With Amkor, Nvidia plans to expand its advanced packaging capabilities in South Korea, with a target increase of over 300% in interposer wafer capacity by the end of 2028. This expansion, combined with the development of new Amkor production sites in the country, is expected to support higher volumes for the group's current and next-generation platforms.
Nvidia also expects its total semiconductor testing capacity to double between calendar year 2026 and 2027, through deployments in the United States, Taiwan, and Singapore. A $1.1 billion capacity reservation advance paid to KYEC is to finance new test sites and equipment in the United States, with high-volume production targeted for April 2027. In Singapore, the expansion of testing activities is scheduled to enter high-volume production in February 2027.
Finally, Nvidia and SPIL plan to increase optical engine manufacturing capabilities in Taiwan for co-integrated optics, to meet connectivity needs of large-scale computing infrastructures. The target is 15x capacity by mid-2027, followed by another doubling targeted for the end of 2028.
