SK Telecom announced it will utilize NVIDIA Blackwell GPUs for AI training and inference, aligning with South Korea’s broader strategy to develop domestic GPU-based clouds to support national AI model development and secure computing resources. This move also reinforces NVIDIA’s partnership with SK Telecom, which has been designated a key collaborator in advancing physical AI initiatives, including projects related to digital twins, robotics, and manufacturing optimization.

NVIDIA: NVIDIA is the primary developer of high-performance computing platforms and AI accelerators, with its Blackwell architecture serving as the latest generation of GPUs tailored for demanding AI and inference tasks. The company supplies the core hardware powering next-generation AI factories and cloud services globally. In this development, NVIDIA’s Blackwell GPUs form the foundation for SK Telecom’s new AI infrastructure deployment.
SK Telecom: SK Telecom is a leading South Korean telecommunications provider that is actively transforming into an AI-focused technology company by developing sovereign AI infrastructure. It is deploying advanced GPU platforms to offer GPU-as-a-service solutions optimized for large-scale AI workloads. In the context of this news, SK Telecom announced plans to use NVIDIA Blackwell GPUs specifically to enable AI training and inference capabilities.

Sovereign AI Expansion: South Korean telecom operators are building domestic GPU-based clouds to support national AI model development and secure computing resources.
Physical AI Partnership: NVIDIA has positioned SK Telecom as a key collaborator in its physical AI initiatives, focusing on digital twins, robotics, and manufacturing optimization using its platforms.