Nvidia Reaffirms 260,000-GPU Supply Commitment to South Korea

Nvidia has reaffirmed plans to supply more than 260,000 graphics processing units to South Korea, keeping one of Asia's largest national AI infrastructure programs on track after the company's leadership change.

Science and ICT Minister Bae Kyung-hoon said the ministry met Nvidia on November 13 to review the supply program and other cooperation projects. Nvidia confirmed the commitments were company-level plans and would continue as announced.

The assurance matters because the October 2025 deal predates Nvidia's recent leadership transition. Practitioners planning around GPU deliveries now have a named confirmation, not just an old press release.

How the government's slice breaks down

The government has secured about 21,000 of the roughly 35,000 GPUs it intends to obtain directly from Nvidia, measured in Nvidia B200 equivalents. The remaining 14,000 are expected to arrive by early 2027.

That's not the whole public allocation. South Korea plans to secure the equivalent of another 15,000 GPUs through the National AI Computing Center, bringing the government's total to more than 50,000 units — roughly one-fifth of the overall 260,000-GPU program.

Put in delivery terms: roughly 60 percent of the direct government procurement is already locked, and the rest lands over the next year or so.

What does this change for Korean companies?

The reaffirmation covers both public-sector procurement and planned deployments by large Korean companies. The original Nvidia statement, issued in October 2025, laid out the private side:

• Samsung Electronics, SK Group and Hyundai Motor Group will each build AI infrastructure using more than 50,000 GPUs
• NAVER Cloud plans to deploy more than 60,000
• The government portion supports sovereign AI projects, cloud providers, research institutes, startups and universities

A Korean government account of the original agreement said the public and private sectors would use the systems for national AI computing, foundation models, industrial applications and physical AI.

For vendors, researchers and startups in the pipeline, the practical takeaway is that access channels — sovereign AI projects, cloud providers and university programs — remain funded and scheduled.

AI factories, not just data centers

Nvidia described the planned buildout as a mix of sovereign cloud capacity and company-operated AI factories. The announced private deployments cover semiconductor manufacturing, automotive systems, robotics, telecommunications and enterprise cloud services.

That spread links the hardware program to several of South Korea's largest technology and industrial groups. It also means the GPU supply touches more than the chip sector — automakers and telecom operators are direct beneficiaries.

Technical cooperation beyond the hardware

Bae said the two sides intend to continue technical cooperation beyond the hardware deliveries. Planned work includes:

• Connecting Korean-designed AI accelerators with Nvidia infrastructure through NVLink
• Developing AI computing capacity
• Conducting physical AI research
• Training specialized workers

The NVLink angle is the one to watch for chip developers. It signals that Korean-designed silicon will plug into Nvidia's rack-scale architecture rather than compete with it head-on.

The ministry will continue monitoring GPU supply and demand across the public and private sectors. It said the government-procured capacity is intended for companies, universities and research institutions, including teams working on South Korea's sovereign foundation-model and AI-for-all initiatives.

A broader Nvidia-Korea push

The reaffirmation comes as South Korea expands several Nvidia-linked infrastructure programs. TNGlobal reported on November 10 that Nvidia, SK Telecom and Korean chip developer Rebellions plan to connect locally designed neural processing units with Nvidia's rack-scale infrastructure for sovereign AI systems targeted for deployment in 2028.

That 2028 target aligns with the NVLink cooperation Bae described, suggesting a coordinated track: hardware deliveries through 2027, then integrated Korean-Nvidia systems the year after. For teams building sovereign AI capacity in Korea, the schedule just got a bit more predictable.