ManufacturingOpinion

NVIDIA’s $390 Billion Supply Bet Faces a Post-Jensen Test

NVIDIA plans to ship its first Vera Rubin-based DSX PODs to U.S. and Central European customers in early 2027, targeting deployment in as little as 14 days at prepared sites.

Answer
NVIDIA’s $390 Billion Supply Bet Faces a Post-Jensen Test

The brief

  1. NVIDIA's supply commitments have increased to $390 billion, primarily due to multiyear memory procurement agreements.
  2. Interim CEO Colette Kress emphasized the importance of supplier relationships and announced expanded agreements with Kinsus, Samsung Electro-Mechanics, and Siliconware Precision Industries.
  3. Kinsus plans to increase substrate capacity for NVIDIA by over 3.5 times, with production expected to start in the second half of 2027.

Deployment

Company
NVIDIA
System
Vera Rubin-based DSX PODs
Stage
Announced
days for deployment
14

Can NVIDIA keep its grip on the supply chain without Jensen Huang?

The brief

  1. NVIDIA's supply commitments have increased to $390 billion, primarily due to multiyear memory procurement agreements.
  2. Interim CEO Colette Kress emphasized the importance of supplier relationships and announced expanded agreements with Kinsus, Samsung Electro-Mechanics, and Siliconware Precision Industries.
  3. Kinsus plans to increase substrate capacity for NVIDIA by over 3.5 times, with production expected to start in the second half of 2027.

Deployment

Company
NVIDIA
System
Vera Rubin-based DSX PODs
Stage
Announced
days for deployment
14

Elon Musk has questioned whether the company can hold on to deals built around its cofounder’s personal relationships. “Without Jensen, at least some of those critical deals will be lost,” he wrote on X, describing arrangements that began with a handshake.

SemiAnalysis founder Dylan Patel raised a similar concern, describing how Huang persuaded even small component suppliers to expand for NVIDIA. “Google’s begging for more capacity from the whole supply chain,” Patel said, pointing to competition for those suppliers’ output.

Huang said he would be willing to return as CEO if the board asked. At the opening of NVIDIA’s latest earnings call, Board Chair Harvey Jones said directors “do not intend to reinstate Jensen as CEO.”

Colette Kress, Interim CEO and Chief Financial Officer, used the call to describe continuing supplier support, including a visit from TSMC leadership that week. “We’ve secured memory. We’ve secured advanced packaging,” she said.

NVIDIA’s supply commitments reached $390 billion, up from $279 billion the previous quarter, with most of the increase tied to multiyear memory procurement. Alongside that broader commitment, new substrate and packaging agreements set out expansion plans extending into 2029.

Putting Capital Behind the Suppliers

“The top priority is going to be our suppliers,” Kress said, sharing that prepayments could help manufacturing partners add capacity faster.

NVIDIA announced expanded agreements with substrate suppliers Kinsus and Samsung Electro-Mechanics, alongside advanced packaging provider Siliconware Precision Industries (SPIL). Substrates provide electrical connections between semiconductor chips and the main board.

Kinsus plans to expand substrate capacity dedicated to NVIDIA by more than 3.5 times current levels at its Youth campus in Taiwan. Production is expected to begin in the second half of 2027, with additional capacity targeted for the third quarter of 2028.

Samsung Electro-Mechanics’ amended agreement focuses on production capabilities at existing factories. SPIL’s Taiwan expansion has multibillion-dollar funding from NVIDIA behind it, with increased contracted packaging capacity due by mid-2029.

Amkor provides an example of how prepayments support that work. NVIDIA’s previously announced agreements, totaling nearly $3 billion, include prepayments for packaging expansion in Arizona and South Korea.

“We want to make sure that we’re there for them so they can move as quickly as possible,” Kress said.

These are multiyear manufacturing plans, with delivery schedules extending beyond the current leadership transition. The announcements leave the financial consequences of changing demand less clear, providing no detailed cancellation terms for the new agreements.

The Customer Has to Be Ready

Expanding production addresses one constraint. NVIDIA also checks whether customers can put the equipment it ships into service.

Asked how the company allocates constrained supply, Kress pointed to “land, power, shell,” referring to the site and physical infrastructure needed for a data center. Available capital is another consideration, while customers outside the investment-grade category also face questions about their committed demand, or offtake.

“We just need to assure that they are ready on their side in terms of being able to accept the compute,” Kress said.

Certain financial commitments to specialized AI cloud providers, known as neoclouds, carry related conditions. NVIDIA says they take effect after providers secure financing and the compute capacity is operational and accepted by NVIDIA. Delays can reduce or eliminate some commitments, a disclosure Kress said clarified existing arrangements.

Vera Rubin’s Next Delivery Milestone

NVIDIA is also moving installation work into manufacturing facilities through DSX POD, its modular AI factory design. Partners integrate and test systems inside containers before shipping them to customers.

The company says deployment can take as little as 14 days at a prepared site, compared with six months for a traditional onsite build. The destination needs its power, cooling and network infrastructure ready to support the equipment.

Kress said first POD shipments to customers in the United States and Central Europe were planned for early 2027. She also reported that hyperscale partners had brought Vera Rubin systems into production in about a month, half the time required for Grace Blackwell.

For enterprise buyers securing AI capacity through cloud providers, that leaves their deployment schedules tied to both equipment availability and the provider’s readiness. NVIDIA’s new agreements document future supply expansion, while the shipment and installation milestones will show how that capacity reaches customers.

Kress closed the call with an assurance about the company’s ability to continue through the transition. “Our customers and our partners are with us. And employees are committed. We’ve got this.”

Written by Amit Raja Naik

Join the discussion

0 answers

People who run these systems answer here. Your verified role is shown, never your employer unless you choose.

Sign in with LinkedIn to answer

Read next

MachineCon USA · from AIM Media House

Where enterprise AI is decided

An invitation-led day for the CIOs, CDOs and Chief AI Officers who own enterprise AI strategy.

  • New York5th edition · 583 Park Avenue
  • Dallas1st edition

Register for MachineCon USASee MachineCon USA →

  • Keynote on the main stage, MachineCon New York 2025
  • AI150 award winners on stage, MachineCon New York 2025
  • At the tables, MachineCon New York 2025
  • Networking, MachineCon New York 2025