Nvidia advances $1.1 billion to KYEC for semiconductor testing expansion
Nvidia said on Wednesday, October 21, 2026, that it has arranged capacity expansions with six suppliers across substrates, advanced packaging, testing and optical engines. For investors and AI buyers, the announcement is a statement of where Nvidia sees the bottleneck and how much of the cost it is prepared to carry.
- Nvidia plans to raise substrate capacity 2.5x over three years, with IBIDEN, Samsung Electro-Mechanics and Unimicron contributing.
- A $1.1 billion capacity reservation advance to KYEC funds new U.S. testing sites, targeting high-volume production in April 2027.
- Watch for volume ramps: Singapore testing in February 2027, KYEC’s U.S. site in April 2027, Unimicron from the second quarter.
- 2.5x substrate capacity over three years; IBIDEN’s Ono Plant alone targets 3x
- $1.1B KYEC advance; IBIDEN’s is only described as multibillion-dollar
- 15x SPIL optical-engine target by mid-2027, then another doubling by end-2028
- >300% Amkor interposer wafer capacity increase by end of 2028
Nvidia announced the plans at GTC Berlin in the release, naming Amkor Technology, IBIDEN, King Yuan Electronics (KYEC), Samsung Electro-Mechanics, Siliconware Precision Industries (SPIL) and Unimicron Technology Corporation. The company says it is “enabling new and existing suppliers” to expand across substrates, advanced packaging, testing and optical engines. It also says the aim is greater resiliency and redundancy in its global supply chain.
Substrates and packaging carry the longest timelines
The substrate target is the broadest figure in the release: 2.5x Nvidia’s capacity over the next three years, with “more underway” beyond the three named agreements. IBIDEN’s agreement was signed and announced earlier this year, and Nvidia is providing a multibillion-dollar capacity reservation advance for the Ono Plant in Japan. That plant targets 3x current capacity, with additional capacity expected to begin ramping in the first quarter of 2028.
Samsung Electro-Mechanics is joining Nvidia’s substrate supply chain with immediate capacity from existing South Korean operations, plus upgrades and new factories in Sejong and Busan. By 2028, Nvidia-allocated output is projected to be “a significant portion” of its total annual package substrate capacity, a share the release does not quantify. Unimicron’s KF substrate expansion in Taiwan, backed by a capacity reservation agreement, should begin adding capacity in the second quarter of 2027 and reach full capacity by mid-2028.
Amkor and Nvidia plan to expand advanced packaging in South Korea, targeting a greater than 300% increase in interposer wafer capacity by the end of 2028. A 300% increase is a fourfold level, and the figure applies to interposer wafers, although the release’s summary bullet describes it as advanced packaging capacity generally.
Testing is the only U.S. element, and it ramps first
Nvidia expects total semiconductor testing capacity to double from calendar 2026 to 2027, supported by buildouts in the U.S., Taiwan and Singapore. The $1.1 billion KYEC advance will support new U.S. testing facilities and equipment, with high-volume production targeted for April 2027, about six months from today. Singapore testing is expected to reach high-volume production in February 2027, about four months out.
SPIL and Nvidia plan to expand Taiwan capacity for the optical engines used in co-packaged optics. The target is a 15x increase by mid-2027, with “aims” for another doubling by the end of 2028. The release does not say whether the doubling is measured against the 15x level or the original baseline.
What the release leaves unanswered
The release gives no baseline capacity for any supplier, so the multiples cannot be converted into units, wafers or systems. It discloses one dollar figure, KYEC’s $1.1 billion, and calls IBIDEN’s advance only “multibillion-dollar.” It gives no amounts for Samsung Electro-Mechanics, Unimicron or Amkor, and no repayment, offset or take-or-pay terms for any advance.
The wording matters for how buyers read the plan. The substrate figure is a plan (“plans to increase”), and the SPIL and Amkor numbers are targets. The release also carries a forward-looking-statements notice listing reliance on third-party manufacturing, assembly, packaging and testing among the risks that could make results differ.
In practice, the advances move supplier capital risk onto Nvidia’s balance sheet in exchange for reserved capacity, a structure the release implies through its terminology but does not detail. Buyers of Nvidia systems get a dated ramp schedule, with testing and optical engines arriving in 2027 and the largest substrate and packaging additions landing in 2028. Despite the heading on U.S. manufacturing, the only U.S. item is KYEC’s testing site; the substrate and packaging expansions are in Japan, South Korea and Taiwan.
“We are in the midst of the largest infrastructure buildout in human history. Demand is extraordinary. Capacity is the constraint,”
Jensen Huang, founder and CEO of NVIDIA, in the company’s release
The BlockWest read. We read the advances as allocation contracts as much as capex support. Nvidia is pre-funding the scarce back-end steps so that rivals have fewer chances to buy the same substrate and packaging slots. The supplier list is wide, but the dollar disclosure is thin, so the real constraint on competitors is the share of each supplier’s output reserved for Nvidia. Investors should treat the undisclosed advance terms as the key variable, because they set how much Nvidia’s cash is tied to supplier execution through 2028.
Next markers: high-volume production in Singapore in February 2027 and at KYEC’s U.S. site in April 2027, plus whether Nvidia’s next quarterly report on its investor relations site quantifies the advances.
BlockWest is a news publication. Nothing here is investment advice. Read our disclaimer and editorial policy.
