SK Telecom has officially launched its GPU-as-a-Service (GPUaaS) offering in its push to expand AI cloud services in South Korea. The new service, hosted at the company’s Gasan AI Data Center (AIDC) in Seoul, was introduced after the completion of pilot operations in late December 2024. This launch is part of SK Telecom’s broader strategy to strengthen its AI infrastructure after announcing its AI infrastructure superhighway initiative at the SK AI Summit 2024 in November.
The Gasan AIDC, which officially opened on December 30, 2024, plays a crucial role in powering this new AI cloud service. With this development, SK Telecom aims to position itself as a key player in AI solutions, providing scalable, high-performance GPU resources to businesses and developers across the region.
SK Telecom’s GPUaaS offering is powered by the high-performance NVIDIA H100 GPUs, which are currently the most advanced GPUs available in South Korea. This service, co-developed with U.S.-based Lambda, allows businesses to select GPU quantities and subscription durations based on their AI service needs.
The service allows for short-term options, such as one- to two-month subscriptions, and enables customization with dedicated servers, firewalls, and private lines to enhance security. Businesses can scale their GPU resources as required, reducing the need for additional infrastructure investments and optimizing their AI workloads.
The Gasan AI Data Center, the core facility supporting SK Telecom’s GPUaaS, is optimized for high-density GPU server operations, featuring a power density of 44kW per rack—almost nine times higher than the national average. The center is a critical component of SK Telecom’s AI Infrastructure SuperHighway strategy, which seeks to establish a nationwide network of AI data centers, GPUaaS offerings, and edge AI solutions. This initiative is designed to boost AI capabilities across South Korea and strengthen the company’s position in the AI infrastructure market.
SK Telecom’s AI Cloud Manager, launched in October 2024, will be used to manage the GPU resources efficiently. This advanced solution allows businesses to treat multiple GPUs as a unified resource, improving performance and optimizing AI training times. By integrating SK Telecom’s GPUaaS with public cloud platforms, businesses can seamlessly blend their existing cloud infrastructure with powerful GPU computing capabilities, further enhancing their AI capabilities and flexibility.
The launch of GPUaaS has already over 100 companies inquiring about the service even before its official debut. SK Telecom’s partnership with Lambda, which also involves investing in stable GPU supply and expertise, has helped the company become a competitive force in the AI cloud market. The service is expected to be pivotal in helping businesses scale their AI workloads, driving innovation across various industries.
SK Telecom’s GPUaaS offering has already attracted significant interest, with more than 100 inquiries from large corporations, universities, and research institutions ahead of its official launch. This reflects the increasing demand for flexible, scalable AI infrastructure solutions across industries.
To celebrate the launch, SK Telecom is offering a 20% promotional discount on subscription prices until the end of February 2025. Additional incentives include a GPU trade-in program and consulting services for cloud cost optimization. The company also plans to roll out Nvidia’s H200 GPUs in the first quarter of 2025, marking SK Telecom as the first in Korea to introduce this advanced technology.
