SK Telecom accelerated its AI infrastructure roadmap with plans to scale its Ulsan artificial intelligence data center (AIDC) to 1 GW and position Korea as a regional AI hub. CEO Jung Jaihun outlined the strategy at the SK AI Summit, highlighting new AIDC hubs across Korea, expanded LNG-powered designs for global markets, and deeper technical partnerships with AWS and NVIDIA. The approach moves SK Telecom beyond telecom-grade cloud toward becoming a full-scope AIDC developer.
The company now anchors three national AIDC regions: Gasan (Seoul), Ulsan (southern), and a new southwestern hub formed under an October MoU with SK Group. SK Telecom aims to replicate the Ulsan model in Southeast Asia, beginning with a Vietnam project that pairs LNG-based power from SK Innovation with cold-energy cooling systems. The 1 GW Ulsan expansion is central to a broader push to attract global capital and scale Korea’s role in Asia’s AI capacity build-out.
SK Telecom also detailed technical co-development tracks with AWS for Edge AI and with NVIDIA for AI-RAN and manufacturing AI use cases. The company plans to deploy more than 2,000 NVIDIA RTX PRO 6000 Blackwell GPUs for a Manufacturing AI Cloud designed to support SK hynix and other industrial units with digital twins and robotics AI workloads. The operator will package these investments into commercial AIDC solution stacks covering power, cooling, clustering, and rack-level optimization.
• Three national AIDC hubs: Gasan (Seoul), Ulsan, and new southwest region
• Ulsan expansion targets 1 GW of AI data-center capacity
• LNG-powered AIDC model to be exported to Southeast Asia, beginning with Vietnam
• Medium-to-long-term R&D partnership with AWS for Edge AI
• AI-RAN development with NVIDIA, government, and academic partners
• 2,000+ NVIDIA RTX PRO 6000 Blackwell GPUs planned for Manufacturing AI Cloud
• Commercial AIDC solution package includes out-rack, clustering, in-rack, and energy systems
• SK hynix and other SK Group manufacturers to consume new AI cloud resources
“We are laying the foundation for South Korea to become Asia’s AI infrastructure hub,” said CEO Jung.
🌐 Analysis
SK Telecom’s shift toward 1 GW-class AIDCs aligns with regional hyperscale trends and Korea’s multi-billion-dollar push to strengthen domestic AI capacity. Its AI-RAN work with NVIDIA echoes similar operator initiatives in Japan, Europe, and the U.S., where cloud-native RAN and edge compute are converging into AI-assisted network architectures. The Manufacturing AI Cloud mirrors moves by global industrial players integrating digital twins and robotics AI into semiconductor and automotive supply chains.
🌐 We’re tracking the latest developments in AI infrastructure. Follow our ongoing coverage at: https://convergedigest.com/category/ai-infrastructure/
