OpenAI partners with Samsung and SK Hynix to boost AI chip supply and data centers in South Korea

Cosmico - OpenAI partners with Samsung and SK Hynix to boost AI chip supply and data centers in South Korea
Credit: OpenAI, Inc./Samsung Group/SK Hynix, Inc.

OpenAI is ramping up its efforts to secure the compute capacity needed to fuel the next generation of artificial intelligence. On Wednesday, the ChatGPT maker announced agreements with Samsung Electronics and SK Hynix, two of the world’s largest memory chip producers, to supply DRAM wafers for the company’s massive Stargate AI infrastructure project and to build new data centers in South Korea.

A High-Level Meeting in Seoul

The deals were formalized in letters of intent following a meeting in Seoul attended by OpenAI CEO Sam Altman, South Korean President Lee Jae-myung, Samsung executive chairman Jay Y. Lee, and SK chairman Chey Tae-won.

Under the agreement, Samsung and SK Hynix plan to scale up production to 900,000 high-bandwidth memory (HBM) DRAM chips per month. According to SK Group, this would more than double the current global industry capacity for HBM — a critical technology for training and running large AI models.

Stargate: OpenAI’s $500 Billion Bet on AI Infrastructure

The supply agreements are part of Stargate, OpenAI’s $500 billion project with Oracle and SoftBank to build next-generation data centers in the U.S. dedicated to AI development. The effort underscores the unprecedented demand for compute power as companies race to train ever-larger models and deploy AI systems at scale.

In just the past month, OpenAI has inked a series of major infrastructure deals:

  • Nvidia committed up to $100 billion to provide OpenAI with access to over 10 GW of compute through its AI training systems.
  • OpenAI partnered with SoftBank and Oracle to construct five Stargate data centers, aiming for a total of 7 GW of capacity.
  • Oracle separately agreed to sell $300 billion worth of compute capacity to OpenAI over five years.

Expanding in South Korea

Beyond chip supply, OpenAI is also deepening its footprint in South Korea. The company is working with the Ministry of Science and ICT to identify sites for AI data centers outside Seoul, while also partnering with SK Telecom to develop a dedicated AI data center. In addition, OpenAI has signed agreements with several Samsung subsidiaries to explore more infrastructure projects in the country.

As part of the collaboration, both Samsung and SK Group will also integrate ChatGPT Enterprise and OpenAI APIs into their operations, extending the partnership beyond hardware supply into enterprise adoption.

The Bigger Picture

The partnerships highlight how OpenAI is aggressively locking in the resources needed to support its ambitious roadmap. With demand for compute capacity outpacing supply and competition intensifying, securing chip production and data center infrastructure has become as strategic as developing the AI models themselves.

For South Korea, the deals also signal the country’s growing importance in the global AI supply chain — not just as a semiconductor hub, but as a key partner in building out the infrastructure of the AI age.

Read more