Nvidia signs billions in supply chain deals to unblock AI chip bottlenecks

CEO Jensen Huang says capacity is the constraint as company targets major expansion in substrates, advanced packaging, and chip testing through 2028

By LineZotpaper AI staff

Published27 October 2026

Read Time2 min

Follow this storyHear when it moves

Illustration: Nvidia signs billions in supply chain deals to unblock AI chip bottlenecks

Nvidia announced a series of multi-billion-dollar agreements with supply chain partners at GTC Berlin on October 21, aimed at overcoming severe bottlenecks in the production of AI accelerators. The deals span substrates, co-packaged optics, advanced packaging, and semiconductor testing, with commitments and production roadmaps stretching through 2028.

Nvidia CEO Jensen Huang described the expansion as a response to extraordinary demand. "We are in the midst of the largest infrastructure buildout in human history. Demand is extraordinary. Capacity is the constraint," he said. "Our supply chain partners are the world's best manufacturers. I'm grateful they are expanding capacity around the globe to scale production and help us deliver the AI supercomputing systems powering this new industrial revolution."

The agreements target two major categories in the AI accelerator supply chain: hardware interconnects and outsourced semiconductor assembly and test (OSAT). On the hardware side, Nvidia signed deals with three of the world's top five substrate manufacturers, Ibiden, Unimicron, and Samsung Electro-Mechanics, to secure a 2.5 times increase in dedicated substrate capacity over three years. This includes supporting facility upgrades and new plants in Japan, Busan, and Sejong. Fubon Research estimates Nvidia will supply around 12.4 million data-center AI GPUs in 2028, each requiring a substrate layer.

For co-packaged optics, Nvidia signed an agreement with Taiwanese manufacturer SPIL to expand optical engine manufacturing capacity by 15 times by mid-2027, with a plan to double again by the end of 2028.

In the OSAT category, Nvidia and Amkor are targeting a 300 percent increase in interposer wafer capacity by the end of 2028 through expanded advanced packaging capacity in South Korea. Amkor confirmed the deal is worth $1.3 billion, bringing Nvidia's total investment in Amkor across its Arizona and South Korea facilities to $3 billion. Separately, Nvidia signed a $1.1 billion deal with semiconductor testing company KYEC to reserve capacity at a new U.S. facility. High-volume production there is slated to begin in April 2027, with the company planning to double its overall testing capacity from 2026 to 2027.

Nvidia's August 2026 financial disclosures showed the company had accumulated $279 billion in outstanding supply and manufacturing-capacity commitments by July 2026. The broader industry is also investing heavily; AMD announced over $10 billion in investments across Taiwan's semiconductor ecosystem in May, specifically targeting advanced packaging and manufacturing.

RSS

The Daily Edition

The front page in your inbox each weekday morning: the lead, the picks, and why each one earned its place.

Analysis

Why This Matters

AI chip supply bottlenecks directly constrain the global buildout of artificial intelligence infrastructure, affecting everything from cloud computing to enterprise AI adoption.
Nvidia's multi-year, multi-billion-dollar commitments are reshaping the geography of semiconductor manufacturing, directing capital to Japan, South Korea, Taiwan, and the United States.
The scale of pre-payment indicates Nvidia expects demand to outstrip supply for years, signaling its determination to maintain market dominance against rising competition.

Background

Nvidia holds a dominant share of the market for AI training and inference chips. Its latest architectures, Blackwell and Rubin, require increasingly complex packaging, including larger substrates and co-packaged optics. The manufacturing base for these components is highly concentrated among a handful of suppliers, making capacity reservation a critical strategic priority. The announcements came from GTC Berlin, a major industry event where Nvidia often unveils supply chain partnerships.

Key Perspectives

Nvidia: Says capacity is the constraint in the largest infrastructure buildout in history and is securing long-term reservations across substrates, advanced packaging, and chip testing to meet extraordinary demand. AMD: Announced over $10 billion in investments in Taiwan's semiconductor ecosystem in May, targeting advanced packaging and manufacturing to compete in the AI chip market.

What to Watch

Whether Nvidia meets its ambitious 2027 targets for 15x co-packaged optics capacity and doubled chip testing capacity.
Geopolitical risks in Taiwan and South Korea, which are central to the supply chains of both Nvidia and its partners.
How AMD and other chipmakers respond to Nvidia's pre-emption of key substrate and packaging capacity.

Sources

Nvidia is bankrolling AI accelerator supply chain buildout to reserve production capacity · Tom's Hardware

Checked

12 claims in this story checked against 1 source: 9 supported.

Show each claim

SupportedNvidia signed multi-billion-dollar supply-chain deals aimed at easing AI-chip bottlenecks.Latest from Tom's Hardware: “Nvidia announced several massive, multi-billion-dollar deals with supply chain partners across substrates, advanced packaging, semiconductor testing, and optical engines to expand production capacity. Aimed at overcoming severe bottlenecks”
SupportedThe company’s expansion plans cover substrates, advanced packaging and chip testing, with supply-chain roadmaps extending through 2028.Latest from Tom's Hardware: “the deals feature commitments and production roadmaps stretching through 2028”
SupportedNvidia announced supply-chain deals at GTC Berlin on October 21 to address bottlenecks.Latest from Tom's Hardware: “On October 21, during GTC Berlin, Nvidia announced several massive, multi-billion-dollar deals with supply chain partners”
SupportedHuang said, “We are in the midst of the largest infrastructure buildout in human history. Demand is extraordinary. Capacity is the constraint.”Latest from Tom's Hardware: ““We are in the midst of the largest infrastructure buildout in human history. Demand is extraordinary. Capacity is the constraint,” said Nvidia CEO Jensen Huang.”
SupportedHuang said, “Our supply chain partners are the world's best manufacturers. I'm grateful they are expanding capacity around the globe to scale production and help us deliver the AI supercomputing systems powering this new industrial revolution.”Latest from Tom's Hardware: ““Our supply chain partners are the world’s best manufacturers. I’m grateful they are expanding capacity around the globe to scale production and help us deliver the AI supercomputing systems powering this new industrial revolution.””
Partly supportedNvidia’s deals with Ibiden, Unimicron and Samsung Electro-Mechanics target 2.5 times dedicated substrate capacity over three years.Latest from Tom's Hardware: “The company says it plans to secure 2.5x dedicated substrate capacity over the next three years.”
Partly supportedThe substrate expansion includes facility upgrades and new plants in Japan, Busan and Sejong.Latest from Tom's Hardware: “Beyond the immediate capacity from Samsung Electro-Mechanics’ existing operations, the deal will support facility upgrades and new plants in Busan and Sejong.”
SupportedFubon Research estimates Nvidia will supply around 12.4 million data-center AI GPUs in 2028, each with a substrate layer.Latest from Tom's Hardware: “Fubon Research estimates that Nvidia would supply around 12.4 million data-center AI GPUs in 2028, up from a projected 8.2 million in 2026. Each one would feature a substrate layer.”
SupportedNvidia’s SPIL agreement targets a 15-fold increase in optical-engine manufacturing capacity by mid-2027, with another doubling planned by the end of 2028.Latest from Tom's Hardware: “The company has signed a deal with Taiwanese optical engine manufacturer SPIL to expand production capacity by 1,400% by mid-2027, with plans to double again by the end of 2028.”
Partly supportedNvidia and Amkor are targeting a 300 percent increase in interposer wafer capacity by the end of 2028.Latest from Tom's Hardware: “Nvidia’s deal with Amkor aims to expand the latter's interposer wafer capacity by more than 300% by the end of 2028”
SupportedAmkor confirmed a $1.3 billion deal, bringing Nvidia’s investment in Amkor’s Arizona and South Korea facilities to $3 billion.Latest from Tom's Hardware: “Amkor confirmed the deal is worth $1.3 billion, bringing Nvidia's total investment in Amkor across its Arizona and South Korea facilities to $3 billion.”
SupportedNvidia signed a $1.1 billion deal with KYEC to reserve capacity at a new U.S. facility, where high-volume production is slated to begin in April 2027.Latest from Tom's Hardware: “On October 13, Nvidia signed a deal to pay KYEC $1.1 billion to reserve testing capacity at KYEC's new $1.4 billion U.S. facility, where high-volume production is slated to begin in April 2027.”