AnswersPublished22 sources
Nvidia Targets 2.5× More AI Chip Substrate Capacity With Six Suppliers
At GTC Berlin on October 21, Nvidia outlined plans with six partners across substrates, packaging, testing and optical engines, targeting 2.5× substrate capacity over three years. The partners are Samsung Electro Mechanics, Ibiden, Unimicron, Amkor, KYEC and SPIL, with roles spanning several stages of manufacturing—...
Published byEdited with GPT-6 LunaImages generated with GPT Image 2
Research answer
Key takeaways
- At GTC Berlin on October 21, Nvidia outlined plans with six partners across substrates, packaging, testing and optical engines, targeting 2.5× substrate capacity over three years.
- The partners are Samsung Electro Mechanics, Ibiden, Unimicron, Amkor, KYEC and SPIL, with roles spanning several stages of manufacturing—not just chip fabrication.
AI Prompt
openai.comCreate a landscape editorial hero image for this Studio Global article: What did Nvidia announce at its October 21 GTC event in Berlin about expanding its AI chip supply chain through Samsung Electro-Mechanics an. Article summary: Nvidia announced at GTC Berlin on October 21, 2026, a six-partner expansion of its AI manufacturing supply chain, spanning substrates, advanced packaging, testing and optical engines; its headline goal was to bring subst. Topic tags: general, general web, news, user generated. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts w
Nvidia’s October 21, 2026, announcement at GTC Berlin was a supply-chain expansion plan, not a single chip-production deal. The company named six partners for work across package substrates, advanced packaging, testing and optical engines, with a goal of increasing substrate capacity for its products by 2.5× over three years. 16
44
What Nvidia announced
The six partners are Samsung Electro-Mechanics, Ibiden, Unimicron Technology, Amkor Technology, King Yuan Electronics (KYEC) and Siliconware Precision Industries (SPIL). Their work spans multiple manufacturing stages, which Nvidia said would help expand capacity and strengthen supply-chain resilience. 16
The substrate target is an aggregate one: Nvidia said agreements with Ibiden, Samsung Electro-Mechanics and Unimicron would contribute to the 2.5× goal. The announcement does not provide individual capacity totals for all three suppliers. 44
What is known about each partner
- Samsung Electro-Mechanics: The company is named as a substrate partner. Reporting says it plans to use existing South Korean capacity, upgrade facilities and build new production lines in Sejong and Busan, with shipments beginning in 2026.
37
47 Samsung Electro-Mechanics is the company named in Nvidia’s announcement; the supplied announcement does not explain its corporate relationship to Samsung Electronics.
- Reported Samsung investment: Reuters reported, citing the Maeil Business Newspaper and industry sources, that Nvidia and Tesla would invest a combined 3.7 trillion won (about $2.8 billion) in FC-BGA substrate production lines in Sejong. The report attributed 2.6 trillion won to Nvidia and 1.1 trillion won to Tesla. These are reported figures, not amounts disclosed in Nvidia’s GTC announcement.
32
- Ibiden: Nvidia’s reported capacity-reservation advance supports expansion of Ibiden’s Ono plant in Japan, with a target of three times its current capacity. Additional output is expected to begin ramping in the first quarter of 2028.
3
10
- Unimicron: It is one of the three substrate suppliers included in the aggregate 2.5× target. The cited announcement materials do not establish an individual expansion schedule or reservation amount.
44
- Amkor: Nvidia and Amkor plan to expand advanced-packaging capacity by more than 300% by the end of 2028. Reporting describes the South Korean expansion specifically in terms of interposer-wafer capacity.
5
49
- KYEC: Nvidia is providing a reported $1.1 billion capacity-reservation advance to support new U.S. testing facilities, with high-volume production targeted for April 2027.
14 A separate report says Nvidia’s testing capacity is expected to double from 2026 to 2027.
1
- SPIL: The reported target for optical-engine capacity is a 15-fold increase by mid-2027, followed by another doubling by the end of 2028. These are future targets, not completed capacity increases.
36
Why the breadth of the plan matters
The announcement covers more than substrate supply. It also addresses packaging, testing and optical-engine capacity—the different manufacturing capabilities Nvidia identified in its partner plans. 16 Jensen Huang summarized the constraint in Nvidia’s announcement as: “Demand is extraordinary. Capacity is the constraint.”
37
That framing helps explain the breadth of the commitments: expanding one component or production stage alone may not resolve limits elsewhere in the manufacturing chain. The 2.5× figure is Nvidia’s aggregate substrate goal; the other supplier figures are separate reported targets and should not be read as equivalent measures.
What remains uncertain
Nvidia’s announcement establishes the six partners, the areas of work and the aggregate substrate target. Some of the more detailed financial amounts and supplier-specific deadlines come from reporting rather than Nvidia’s announcement. In particular, the Samsung investment split is attributed to industry sources, while the supplied materials do not verify every individual plant schedule or the proportion of Samsung Electro-Mechanics’ capacity that Nvidia may use. 16
32
There is also a business risk to consider, rather than a confirmed outcome: if demand for AI infrastructure weakens, newly expanded or dedicated facilities could be harder for suppliers to keep fully utilized. The available information does not specify the contract terms needed to assess who would bear that risk. Overall, Nvidia’s announcement signals an effort to secure capacity across several supply-chain stages, but the headline targets remain plans to be delivered over time.
Studio Global AI
Continue your research
This page includes a source-backed answer you can continue inside Studio Global.
People also ask
What is the short answer to "Nvidia Targets 2.5× More AI Chip Substrate Capacity With Six Suppliers"?
At GTC Berlin on October 21, Nvidia outlined plans with six partners across substrates, packaging, testing and optical engines, targeting 2.5× substrate capacity over three years.
What are the key points to validate first?
At GTC Berlin on October 21, Nvidia outlined plans with six partners across substrates, packaging, testing and optical engines, targeting 2.5× substrate capacity over three years. The partners are Samsung Electro Mechanics, Ibiden, Unimicron, Amkor, KYEC and SPIL, with roles spanning several stages of manufacturing—not just chip fabrication.