Sam Altman is heading back to South Korea. The OpenAI CEO is scheduled to visit Samsung Electronics on June 15 for the “2026 Device eXperience (DX) Insight Talk” at the company’s Suwon campus, where the conversation will center on AI adoption and deeper collaboration between the two tech giants.
To understand why Altman keeps booking flights to Seoul, you need to understand Stargate. That’s OpenAI’s plan to build a global network of hyperscale AI data centers by 2029, representing the company’s largest infrastructure endeavor to date. The project carries a reported price tag of $500B.
AI data centers are extraordinarily hungry for one thing above all else: memory chips. Specifically, DRAM and high-bandwidth memory (HBM), the silicon that lets AI models process massive datasets at speed. Samsung happens to be one of the world’s largest producers of both.
The groundwork for this partnership was laid well before the June visit. In October 2025, Altman met with Samsung leadership in a high-profile gathering that also included South Korean President Lee Jae-myung. That meeting produced letters of intent for Samsung and fellow Korean chipmaker SK Hynix to supply OpenAI with up to 900,000 DRAM wafers per month, representing approximately 40% of the world’s total production capacity for those chips.
