Nvidia Advances $1.1B to KYEC for New US Chip Test Sites

Nvidia said on 21 October 2026, at GTC Berlin, that it is giving King Yuan Electronics (KYEC) a $1.1 billion capacity reservation advance to support new U.S. chip testing facilities and equipment, with high-volume production targeted for April 2027. The advance is one of the commitments with six suppliers named in the same release, covering substrates, advanced packaging, testing and optical engines.

In plain terms, a capacity reservation advance is money paid up front to a supplier to build or reserve production capacity for the buyer. Nvidia’s release frames the whole package in one line from Jensen Huang: “Demand is extraordinary. Capacity is the constraint.”

Business Pill · CONCENTRATION

A Business Engineer explainer on concentration. It covers the concept in general and says nothing about Nvidia or the suppliers named.

Watch · The Keynote

NVIDIA GTC Berlin 2026 keynote with CEO Jensen Huang, 21 October, on NVIDIA’s YouTube channel. Quotations in this piece come from Nvidia’s release, not from the spoken keynote.

What Nvidia Announced

The release says Nvidia is working with Amkor Technology, IBIDEN, King Yuan Electronics (KYEC), Samsung Electro-Mechanics, Siliconware Precision Industries (SPIL) and Unimicron Technology Corporation to expand manufacturing “across substrates, advanced packaging, testing and optical engines.” Nvidia says the aim is “strengthening the resiliency and redundancy of NVIDIA’s global supply chain.”

“We are in the midst of the largest infrastructure buildout in human history. Demand is extraordinary. Capacity is the constraint,” Huang, Nvidia’s founder and CEO, said in the release. “Our supply chain partners are the world’s best manufacturers.”

What Was Posted on X

Wall St Engine lists the agreements point by point, leading on Amkor’s plan to raise advanced packaging interposer wafer capacity in South Korea by more than 300% by the end of 2028, and repeats Huang’s line that capacity is the constraint. The figures in this piece come from Nvidia’s release itself.

The Supplier Deals, One by One

A short glossary first. A substrate is the base board a chip is mounted on. Advanced packaging joins a graphics processor and its memory into one part, and an interposer is the layer that connects them. Testing checks each chip before it ships, and an optical engine turns electrical signals into light for co-packaged optics, where the optics sit next to the chip.

• Substrates. Nvidia plans to increase substrate capacity for its products by 2.5x over the next three years, the release says, through agreements with IBIDEN, Samsung Electro-Mechanics and Unimicron, “with more underway.”
• IBIDEN. Under an agreement IBIDEN signed and announced earlier this year, Nvidia is providing a “multibillion-dollar capacity reservation advance” for IBIDEN’s Ono Plant in Japan, where the expansion targets capacity reaching 3x current levels, with additional capacity expected to begin ramping in the first quarter of 2028.
• Samsung Electro-Mechanics. Joining Nvidia’s substrate supply chain with capacity from its existing operations in South Korea, plus upgrades and new factories in Sejong and Busan. By 2028, the release says, production allocated to Nvidia is projected to account for “a significant portion” of the company’s total annual package substrate capacity.
• Unimicron. Nvidia is supporting its KF substrate expansion in Taiwan with a capacity reservation agreement; the expansion is expected to begin adding capacity in the second quarter of 2027 and reach full capacity by mid-2028.
• Amkor. Nvidia and Amkor plan a “greater than 300% increase in interposer wafer capacity by the end of 2028” in South Korea, alongside new Amkor facilities there.
• KYEC and testing. Nvidia expects its total semiconductor testing capacity to double from calendar 2026 to 2027, with buildouts in the U.S., Taiwan and Singapore. The $1.1 billion advance supports KYEC’s new U.S. facilities, and expanded testing operations in Singapore are expected to begin high-volume production in February 2027.
• SPIL. Nvidia and SPIL plan to expand optical-engine manufacturing in Taiwan, targeting an increase of 15x by mid-2027, “with aims for another doubling by the end of 2028.”

Why the Money Goes Up Front

Two of the six supplier commitments carry a stated amount of money: the $1.1 billion advance to KYEC and the multibillion-dollar advance to IBIDEN. Unimicron’s is described as a capacity reservation agreement, without a figure. As we read it, the IBIDEN and KYEC advances put Nvidia money into substrates and testing ahead of the capacity coming online, Unimicron’s is a reservation agreement with no amount given, and the Amkor and SPIL items are described as plans to expand. The release’s own summary of why is Huang’s line: “Capacity is the constraint.”

Connecting the Coverage

This is one of several releases Nvidia put out at GTC Berlin on 21 October. Our DSX POD piece reported Nvidia’s claim that factory-built AI factory units can cut deployment at a prepared site to as little as 14 days, and our Vera Rubin piece reported its release figure of “30x higher throughput per megawatt” for agentic workloads. Earlier the same day, our report on Nvidia’s data centre backing covered The Information’s finding that Nvidia’s guarantees and other commitments supporting new data centers reach up to $164.5 billion.

On 20 October, our Supermicro piece reported that the server maker now estimates first fiscal quarter revenue of $18.5 billion to $19.5 billion, citing demand for AI platforms. A separate analysis on this site looked at how Nvidia’s newer tray designs cut assembly time. As we read it, what this release adds is the supplier layer: substrates, packaging, testing and optics, the steps between a finished chip design and a built system.

The Structural Read

As we read it, the money matters because it is paid up front: the release describes both the KYEC and IBIDEN deals as capacity reservation advances, so Nvidia carries part of the cost of building supplier capacity before that capacity comes online.

As we read it, the spread of locations is part of the message: the release names Japan, South Korea, Taiwan, Singapore and the U.S., and says the goal is “strengthening the resiliency and redundancy” of Nvidia’s supply chain.

As we read it, the dates in the release set out a timetable: Singapore testing in February 2027, KYEC in April 2027, SPIL’s 15x by mid-2027, Unimicron at full capacity by mid-2028 and Amkor’s target by the end of 2028.

Three Implications

FOR CHIP PACKAGING AND TESTING SUPPLIERS The release names six suppliers, Amkor, IBIDEN, KYEC, Samsung Electro-Mechanics, SPIL and Unimicron, and says more substrate agreements are underway.

FOR COMPANIES WAITING ON AI SYSTEMS Nvidia says it expects total testing capacity to double from 2026 to 2027 and plans to increase substrate capacity 2.5x over three years.

FOR U.S. MANUFACTURING The release says the $1.1 billion KYEC advance supports new U.S. testing facilities and equipment, with high-volume production targeted for April 2027.

The Business Engineer Lens

This story maps onto the Business Engineer framework AI’s Geopolitical Chokepoint.

The analysis opens on the physical layer of AI: “The digital economy rests on a physical foundation that most people in that economy never see.”

As we read it, Nvidia’s release is about one such layer, the substrates, packaging, testing and optics between chip design and finished systems. The analysis was published in December 2025 and is about critical materials such as rare earths; it does not discuss these suppliers or this release.

What Is Not Established

The release, as we read it on 21 October, gives a dollar figure only for KYEC. It does not give the size of IBIDEN’s “multibillion-dollar” advance, the value of the other agreements, the U.S. location of KYEC’s new facilities, or the name of the company behind the expanded testing in Singapore. Several of the targets reach into 2027 and 2028, so they are plans, not finished capacity. Drawing Amkor’s “greater than 300%” increase as 4x in our poster is our arithmetic.

The Bottom Line

Nvidia used GTC Berlin to announce a $1.1 billion capacity reservation advance to KYEC for new U.S. chip testing facilities, alongside plans with five other suppliers across substrates, packaging and optical engines. As we read it, the release puts dates and multiples on Huang’s line that “Capacity is the constraint,” with most of the targets landing between February 2027 and the end of 2028.