NVIDIA and SK Telecom to build gigawatt-scale Korean AI Cloud on DSX

NVIDIA and SK Telecom unveiled plans for a gigawatt-scale AI Cloud in Korea built on NVIDIA's DSX platform, with the first 'AI factory' slated to come online in 2027. The deal extends NVIDIA's sovereign-AI infrastructure strategy — partnering with national telecoms and governments to stand up large, country-specific compute footprints — into one of Asia's most advanced tech economies.

The announcement came amid a broader NVIDIA showcase at GTC Taipei, where Jensen Huang introduced the Vera Rubin computing platform and Vera CPU, a Microsoft PC collaboration, the Alpha Mini 2 open model for self-driving cars, and Blackwell RTX GPUs. The Korean build complements NVIDIA's existing dominance — the company holds roughly 90% of AI accelerator sales — by locking in long-term demand for its hardware.

The gigawatt framing signals the scale at which AI infrastructure is now planned: power, not just chips, is the constraint. For SK Telecom, it's a bid to become a regional AI-cloud provider; for NVIDIA, it's another anchor tenant in a global network of AI factories. The 2027 timeline is the caveat — these are multi-year commitments subject to power availability, financing, and the pace of chip supply, and 'announced plans' for gigawatt facilities have a history of slipping.