NVIDIA announced on Oct. 21 at its GTC Berlin conference that it is backing six supply-chain partners as demand grows for its accelerated-computing and AI infrastructure platforms. The partners are Amkor Technology, IBIDEN, King Yuan Electronics, Samsung Electro-Mechanics, Siliconware Precision Industries and Unimicron Technology. Their work spans package substrates, advanced packaging, semiconductor testing and optical engines.
IBIDEN is expanding its Ono plant in Japan toward three times its current capacity, with additional output expected to begin ramping in the first quarter of 2028. Unimicron’s KF substrate expansion in Taiwan is expected to begin adding capacity in the second quarter of 2027 and reach full capacity by mid-2028.
NVIDIA and Amkor are targeting an increase of more than 300% in interposer-wafer capacity in South Korea by the end of 2028, alongside new manufacturing facilities there. NVIDIA also expects total semiconductor-testing capacity to double between calendar 2026 and 2027 through projects in the United States, Taiwan and Singapore. Testing facilities and equipment, with high-volume production targeted for April 2027. Expanded testing operations in Singapore are expected to begin high-volume production in February 2027.
SPIL is expanding in Taiwan to produce optical engines used in co-packaged optics, a technology that places optical connections close to computing components. The company is targeting a 15-fold increase in optical-engine capacity by mid-2027 and aims to double that capacity again by the end of 2028. Across the plans, most added output is scheduled for 2027 or 2028, so the announcement represents capacity commitments and construction or expansion targets rather than completed production. NVIDIA chief executive Jensen Huang described capacity as the constraint on the company’s ability to meet demand, while the company characterized the agreements as a way to add redundancy and resilience across its global supply chain.