Hlib Chabaniuk

NVIDIA expands supply chain commitments to support AI infrastructure growth

NVIDIA expands supply chain commitments to support AI infrastructure growth
NVIDIA boosts AI supply

NVIDIA is broadening manufacturing and supply agreements to match rising demand for its accelerated computing and AI infrastructure platforms. The expansion spans substrates, advanced packaging, wafer production and contract manufacturing across Asia, the U.S. and Europe, with new capacity additions extending into 2028.

Highlights

  • NVIDIA announced multiyear agreements with suppliers like Amkor, Foxconn, Samsung Electro-Mechanics, SPIL and TSMC to significantly expand capacity across substrates, packaging, and system production for current and next-generation AI platforms.
  • Kinsus will increase ABF substrate manufacturing capacity for NVIDIA by more than 3.5 times, with production starting in the second half of 2027 and added capacity online in Q3 2028; Amkor's advanced packaging expansion is supported through nearly $3 billion in agreements.
  • NVIDIA expects third quarter 2024 to mark the first volume shipments of Vera Rubin NVL72 and Vera CPUs using TSMC's 5nm, 3nm, and CoWoS technologies, amid a doubling of semiconductor testing capacity from 2026 to 2027 and a 15 times increase in SPIL's optical-engine capacity by mid-2027.

Capacity expansion plans across substrates and packaging

As reported by IPRsoftwaremedia, citing NVIDIA Newsroom, the company is enlarging commitments with Amkor Technology, Foxconn, Kinsus Interconnect Technology, Samsung Electro-Mechanics, Siliconware Precision Industries and TSMC to support long-term production growth for current and next-generation platforms.

Debora Shoquist, executive vice president of operations at NVIDIA, says the company views its supply chain as a strategic advantage and that the latest agreements expand capacity across components, advanced packaging and system production needed by customers.

Under a new agreement, Kinsus plans to raise advanced ABF package substrate manufacturing capacity dedicated to NVIDIA by more than 3.5 times current levels at its Youth campus in Taoyuan, Taiwan. Production is expected to begin in the second half of 2027, with the added capacity targeted for the third quarter of 2028.

NVIDIA and Samsung Electro-Mechanics have also amended their existing substrate supply agreement to improve operational capabilities at current factories and support higher production of NVIDIA accelerated computing platforms.

The company says its advanced packaging plans include a new expansion with SPIL as well as previously announced agreements backing Amkor's manufacturing buildout in Arizona and South Korea. NVIDIA is supporting Amkor's advanced packaging expansion through multiyear agreements totaling nearly $3 billion, including prepayments intended to help scale capacity for long-term production needs.

NVIDIA's advanced packaging operations are also expanding in Taiwan under a multiyear agreement backed by multibillion-dollar funding from the company. The project includes new production lines covering bumping, interposer wafer fabrication, assembly and testing for semiconductor products with high requirements for computing performance, bandwidth, power efficiency and system integration.

Industry impact and production ramp

The company says the third quarter marks the first volume shipment of NVIDIA Vera Rubin NVL72 and Vera CPUs for agentic AI, products that use TSMC's 5nm, 3nm and CoWoS advanced packaging technologies. Foxconn is also deepening work with NVIDIA on next-generation GPU and networking platforms through its global production network across Asia, North America and Europe.

These agreements add to earlier manufacturing expansion commitments with Amkor, IBIDEN, King Yuan Electronics, Samsung Electro-Mechanics, SPIL and Unimicron. NVIDIA says those efforts include an aggregate 2.5 times increase in substrate capacity over three years, broader advanced packaging, a doubling of semiconductor testing capacity from 2026 to 2027 and a 15 times increase in SPIL's optical-engine capacity by mid-2027, with a further doubling targeted by the end of 2028.

For the semiconductor supply chain, the expanded commitments indicate continued investment in capacity for AI systems as demand shifts from chips alone to broader infrastructure, including packaging, testing and system-level manufacturing. The plan also reinforces Taiwan, South Korea and the U.S. as key production bases in NVIDIA's long-term manufacturing strategy.

We have also analyzed SpaceX debt sale plans to finance Nvidia AI chips through an off-balance-sheet structure.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.