Hyundai Motor to invest $6.3b in AI, robot factory in S Korea

Hyundai Motor Group and the South Korean government on February 27, signed a deal to invest about 9 trillion won (US$6.3 billion), the South Korean land ministry said.

Hyundai Motor Group will invest about 5.8 trillion won (US$4.04 billion) in an AI data centre deploying 50,000 graphics processing units, the land ministry said.

The group will spend 400 billion won (US$278.8 million) on a factory to produce robots, including wearable robots, the land ministry said.

It also plans to invest 1 trillion won (US$697.2 million) in hydrogen production facilities and 1.3 trillion won (US$906.6 million) in solar energy projects, the land ministry said.

These investments support the development of the Saemangeum land reclamation area, a project started over 20 years ago to improve farmland and industrial growth in the Jeolla provinces.

🔗 Source: Reuters

🧠 Food for thought

Implications, context, and why it matters.

Nvidia’s Blackwell GPUs are at the heart of South Korea’s AI ambitions

Hyundai Motor Group expects the 50,000 GPUs for its Saemangeum AI data center to be NVIDIA chips based on the Blackwell architecture 1.
The buildout sits within a national drive that totals more than a quarter-million NVIDIA GPUs across sovereign clouds and AI factories, while Samsung and SK Group each pursue 50,000+ GPU AI factory efforts 2.
Hyundai Motor Group aims for an end-to-end “physical AI” stack for robotics and autonomous systems, using NVIDIA Omniverse Enterprise (a software platform for simulating and operating industrial environments) to create factory digital twins for simulation and testing 3.
The Saemangeum schedule sets construction to begin in 2027 and target completion in 2029, according to Hyundai Motor Group 1.

Hyundai’s AI investment signals a push toward deeper in-house infrastructure

In Saemangeum, Hyundai Motor Group pairs the AI data center with a robotics production base plus hydrogen production facilities and solar energy projects to tighten control over its industrial setup 1.
The approach answers a concern about becoming “an assembler of someone else’s AI,” keeping more of the AI pipeline in-house from R&D through deployment in vehicles and on factory floors 4.
The large scale matches NVIDIA CEO Jensen Huang’s view of “intelligence as a new export” tied to Korea’s AI factories 2.
Power supply remains central, with up to 10 gigawatts of renewable capacity planned in the Saemangeum area by 2030 to draw energy-hungry sites such as AI data centers 5.