Skip to main content

NVIDIA expands substrate and packaging supply agreements

19 NOVEMBER 2026·2 MIN READ·4 SOURCES

NVIDIA said it has agreements with six manufacturing partners covering substrates, advanced packaging, wafers and contract manufacturing. The commitments are intended to align production capacity with demand for its accelerated-computing and AI-infrastructure platforms.

Key takeaways · 4

  • 01

    Kinsus plans to expand NVIDIA-dedicated advanced ABF substrate capacity to more than 3.5 times its current level at its Taoyuan campus.

  • 02

    Kinsus expects production at the expanded campus to begin in the second half of 2027; additional capacity is targeted for the third quarter of 2028.

  • 03

    SPIL and NVIDIA aim to increase advanced-packaging capacity contracted specifically for NVIDIA by mid-2029.

  • 04

    TechTarget reported that most of the recent increase in NVIDIA’s supply commitments came from multiyear memory procurement.

Agreements span six partners

NVIDIA said its agreements involve Kinsus, Samsung Electro-Mechanics, SPIL, Amkor, Foxconn and TSMC, across manufacturing inputs that include substrates, advanced packaging, wafers and contract manufacturing.[1] The company described its supply chain as a strategic advantage.[1] It said the partner commitments align supply and manufacturing capacity with demand for its accelerated-computing and AI-infrastructure platforms.[1] The announcement therefore covers several parts of the production chain, rather than a single supplier or manufacturing input.[1]

Kinsus plans a large campus expansion

Kinsus plans to expand NVIDIA-dedicated advanced ABF package-substrate capacity at its Youth campus in Taoyuan, Taiwan, to more than 3.5 times its current level.[1] It expects production at the expanded campus to begin in the second half of 2027.[1] NVIDIA says the additional capacity is targeted for the third quarter of 2028.[1] Kinsus said NVIDIA is investing more than US$1 billion to support the campus expansion and secure long-term advanced-substrate supply for next-generation AI platforms.[2] Kinsus also said its supplier relationship with NVIDIA has lasted more than a decade.[1]

Samsung and SPIL have separate plans

NVIDIA and Samsung Electro-Mechanics amended their existing substrate supply agreement to improve operational capabilities at existing factories.[1] The collaboration covers substrate manufacturing in Sejong and Busan, South Korea, for NVIDIA compute and high-speed networking technologies.[1] Separately, NVIDIA and SPIL agreed to expand advanced-packaging operations in Taiwan under a multiyear agreement supported by multibillion-dollar NVIDIA funding.[1] Their planned SPIL production lines will provide turnkey bumping, interposer wafer fabrication, assembly and testing.[1] The companies aim to increase advanced-packaging capacity contracted specifically for NVIDIA by mid-2029.[1]

Commitments build on earlier supply plans

SPIL said its multiyear expansion will triple its advanced-packaging capacity.[3] NVIDIA said agreements announced in July and October 2026 total nearly $3 billion and include prepayments.[1] In October, NVIDIA described plans for a 2.5-fold aggregate increase in substrate capacity over three years, doubled semiconductor testing capacity from 2026 to 2027, and a 15-fold increase in SPIL optical-engine capacity by mid-2027.[1] TechTarget reported that NVIDIA’s supply commitments reached $390 billion at the end of the third quarter, up from $279 billion three months earlier, with most of the increase coming from multiyear memory procurement.[4] Analyst Stephen Sopko said memory, substrates and packaging require substantial capacity and have long lead times.[4]

For teams planning AI infrastructure or semiconductor supply, the announcement offers supplier-specific expansion plans and target dates to distinguish from general commitments. Those dates can inform scenario planning, but they are not interchangeable across suppliers or types of capacity.

Why it matters
For teams

AI training for your team

kju.ai is built for teams: daily sessions, team analytics, and rollout support for your organization.

How this developed

  1. 19 November 2026

    NVIDIA expands substrate and packaging supply agreements

Sources

AI fluency, one session a day, built for your work.