Overview
- Nvidia announced a string of partnerships on Monday that include a multi‑year memory co‑development pact with SK hynix and joint plans with SK Telecom, Naver and Doosan to build large AI data centres.
- SK Telecom and Nvidia committed to a gigawatt‑scale AI cloud project with the first 'AI factory' targeted to begin operations in 2027, and Naver said it will start with a 55‑megawatt expansion in Sejong before scaling toward gigawatt capacity.
- The SK hynix deal focuses on co‑developing next‑generation high‑bandwidth memory for Nvidia’s top servers, aiming to steady supply for platforms such as Vera Rubin that require terabytes of HBM4 per system.
- Nvidia also confirmed technology collaborations with LG on humanoid robotics and future data‑center design, and said talks with Hyundai and the Korean government about a local R&D center and wider Saemangeum plans are advancing.
- By aligning memory roadmaps, data‑centre builds and robotics work with major Korean firms, the deals are meant to shorten lead times for AI infrastructure and expand domestic capacity for sovereign, physical and personal AI services.