Samsung leadership meets Sam Altman on memory supply and Stargate cooperation

According to Chosun Ilbo, the session brought together about 50 leaders from Samsung affiliates with Altman. It had two tracks.

**Hardware.** Samsung is one of the few suppliers of high-bandwidth memory, which constrains frontier AI buildouts. Stargate, OpenAI's multi-site data center program, needs enormous HBM and DRAM volumes. Locking in supply and co-development with Samsung is strategic for OpenAI amid tight HBM4 markets. NVIDIA itself just trimmed Vera Rubin HBM4 bandwidth targets.

**Enterprise adoption.** Executives discussed building decision pipelines and structuring private corporate data with generative models. That aligns with a broader shift: days earlier, Samsung launched a group-wide 'AI transformation' rolling out Gemini, ChatGPT and Claude across software, marketing, R&D and manufacturing. This reverses the 2023 restrictions imposed after engineers leaked source code to ChatGPT; Samsung previously relied on its in-house Gauss model.

The meeting shows how tightly AI labs and memory makers are intertwined: OpenAI is both a major customer of Samsung components and a software vendor to Samsung's workforce. SK Hynix and Micron compete for the same Stargate commitments, so a high-profile Samsung alignment is a negotiating signal.

Caveats: no contract figures were disclosed, and 'alignment sessions' can be ceremonial. Watch for:

- Formal supply agreements or Korean data center announcements.

- How Samsung balances OpenAI ties with its Google partnership, as its Gemini-powered Android XR glasses launch in mid-December.