SK Telecom (017670.KS) shares are surging more than 10% after Nvidia CEO Jensen Huang introduced the company as a partner in the physical artificial intelligence (AI) sector, drawing rapid buying interest.
According to the Korea Exchange on Tuesday, SK Telecom shares were trading at 114,100 won, up 13,500 won or 13.42% from the previous session. The stock had hovered in the low 100,000 won range until 2 p.m. before rallying sharply.
SK Telecom was introduced as a key partner in the manufacturing and physical AI sector during Huang's GTC Taipei keynote, held in Taiwan on Tuesday local time. The keynote video showcased SK Telecom's digital twin technology applied to SK hynix's semiconductor manufacturing process using Nvidia's Omniverse.
GTC is an AI and GPU conference hosted by Nvidia, and this GTC Taipei was held in conjunction with Computex, the global IT exhibition. SK Telecom's appearance in Huang's keynote marks its second after GTC 2026 held in San Jose, U.S., in March.
At that time, SK Telecom was introduced as a strategic partner of Nvidia alongside global robotics and AI companies including Techman Robot and Boston Dynamics, presenting its factory digital twin technology. SK hynix completed a proof of concept (PoC) with SK Telecom last year on applying digital twin technology to semiconductor fabs, targeting the construction of an autonomous factory by 2030, and plans to pursue phased commercialization going forward.
Digital twin is a core technology of physical AI that replicates actual factories and facilities in virtual space, allowing pre-verification of the impact of process changes or equipment layouts through simulation. SK Telecom has developed an agentic digital twin modeling technology that automates and intelligently processes manufacturing site data in virtual environments using Nvidia's agent toolkit.
The company explained that this has improved efficiency in the digital twin construction and operation process, including data conversion and scene optimization. SK Telecom is currently advancing its platform by integrating the Nvidia Omniverse library to improve the loading speed of large-scale 3D scenes and the efficiency of GPU and memory usage.
Based on this, the company plans to expand its public and enterprise businesses as a full-stack AI operator equipped with solutions across the entire spectrum, from AI infrastructure to models and services.
