TodayToday's edition
NVIDIA announced capacity deals with six suppliers covering substrates, interposers, testing and co-packaged optics, including a $1.1 billion advance to KYEC for US test sites.
1 min read
Graphic from NVIDIA's release naming its six supply-chain partners; it is a logo card, not a photo of the facilities.
Image: NVIDIA
Source · Press image from the NVIDIA newsroom, used for news reporting.
NVIDIA said on October 21 at GTC Berlin that it is working with six suppliers to expand manufacturing capacity for substrates, advanced packaging, chip testing and co-packaged optics. In its press release, the company said substrate agreements with IBIDEN, Samsung Electro-Mechanics and Unimicron will together drive a 2.5x increase in substrate capacity over the next three years.
Optical-engine capacity carries the largest stated target, 15x by mid-2027, while substrates are planned at 2.5x.
Illustration (AI-generated)
Samsung Electro-Mechanics is joining NVIDIA's substrate supply chain, with immediate capacity from existing operations in South Korea and facility upgrades and new factories planned in Sejong and Busan, NVIDIA said. By 2028, the production allocated to NVIDIA is projected to be a significant portion of the company's total annual package substrate capacity. NVIDIA is providing IBIDEN a multibillion-dollar capacity reservation advance for its Ono Plant in Japan, which targets 3x current capacity, with more ramping from the first quarter of 2028. Unimicron's KF substrate expansion in Taiwan is due to start adding capacity in the second quarter of 2027 and reach full capacity by mid-2028.
On packaging, NVIDIA and Amkor plan to raise interposer wafer capacity in South Korea by more than 300% by the end of 2028. On testing, NVIDIA expects its total semiconductor testing capacity to double from 2026 to 2027. A $1.1 billion capacity reservation advance to KYEC will fund new US test facilities, with high-volume production targeted for April 2027; Singapore testing is due to start high-volume production in February 2027. SPIL aims to expand optical-engine capacity for co-packaged optics 15x by mid-2027, with another doubling targeted by the end of 2028.
For AI buildouts, the constraint is moving from wafers to the parts around them: substrates, interposers, test time and optics. NVIDIA chief executive Jensen Huang said in the release that demand is extraordinary and capacity is the constraint.
NVIDIA did not give dollar amounts for the Samsung Electro-Mechanics, Unimicron or Amkor agreements, nor chip volumes the new capacity would support. The multiples are targets, and the release carries NVIDIA's standard forward-looking-statement caveats.
