SK Telecom has introduced 'SKT GPU-as-a-Service' (SKT GPUaaS) at its AI Data Center (AIDC) in Gasan, Seoul, South Korea.
The company has invested in Lambda, a global GPU cloud company, to ensure a stable supply of GPUs and expertise for this on-demand artificial intelligence (AI) cloud service.
Latest: SK Telecom Expands AI Solutions with New Partnership
SKT GPUaaS allows businesses to select the number and duration of GPUs required based on the scale or purpose of AI services. Customized packages, such as standalone servers, firewalls, and dedicated lines, can be configured. Prices are flexible and dependent upon the contract period, number of GPUs, and billing method.
By using the 'AI Cloud Manager' launched in October, GPU resources can be managed more efficiently. The AI Cloud Manager treats multiple GPU resources as a single computer to maximize GPU performance and reduce the learning time needed for AI development.
Also Read: SK Telecom Invests in AI Video Startup, Twelve Labs
The GPUaaS service is powered by NVIDIA H100 Tensor Core GPUs. SK Telecom plans to introduce the latest NVIDIA H200 Tensor Core GPUs in the first quarter of 2025. By bringing these capabilities to South Korea, SK Telecom aims to accelerate the development of AI technology and services for local companies and expand its GPUaaS customer base.
Kim Myong-gook, Head of SK Telecom's GPUaaS business, stated, “The launch of GPUaaS is the first step in actively engaging customers in our AI Data Center business. This move is essential in establishing SK Telecom as a key player in the AI infrastructure sector.”
“Our objective is to enhance our clients’ competitiveness by using our GPU as a platform for their AI services,” he concluded.
More from SK Telecom:
CelcomDigi and SK Telecom Forge AI Partnership with New MoU
SK Telecom, Nokia Unveil AI-Driven Fiber Sensing for Enhanced Network Reliability
SK Telecom Unveils Autonomous Robots Utilizing Telco Edge AI
