NVIDIA and Korea Widen Sovereign AI Partnership Across Chips, Models and Security

The expanded collaboration pairs national AI infrastructure with inference software, agent security tools and faster semiconductor simulation.

NVIDIA and Korea's Ministry of Science and ICT are widening their sovereign AI partnership, covering chips, data centers, talent, startups, models, agents and security. Korean NPU and XPU makers may now explore joining the NVLink Fusion ecosystem, and the plan adds training, certification, hackathons and seminars aimed at better GPU use. Under the national AI for All initiative, NVIDIA will work with three consortia and four Korean foundation model builders, using Dynamo software to speed up and scale inference. Security is also moving forward: AIM Intelligence is wiring its Starfort platform into OpenShell, a sandboxed runtime with policy controls for autonomous agents, while LG AI Research has opened API access to K-EXAONE-2.5, which was built on Nemotron datasets including a 7 million persona Korean synthetic dataset. Separately, Samsung is using NVIDIA tools to cut semiconductor simulation times sharply, with one thermal analysis reported at roughly 1,600x faster.