SK Telecom has officially launched its GPU-as-a-Service (GPUaaS) platform, hosted at its AI Data Center (AIDC) in Gasan, Seoul, offering scalable and flexible on-demand AI cloud services. This initiative, backed by SKT’s investment in Lambda, a global GPU cloud company, enables the company to ensure stable GPU supply and technical expertise. The GPUaaS platform provides businesses the flexibility to choose GPU resources based on their AI service requirements, with customizable packages including standalone servers, firewalls, and dedicated network lines.
The service is powered by NVIDIA H100 Tensor Core GPUs, with plans to introduce the NVIDIA H200 Tensor Core GPUs in Q1 2025, reinforcing South Korea’s AI infrastructure capabilities. AI Cloud Manager, launched in October, supports the platform by managing multiple GPUs as a unified computing resource, optimizing performance and reducing AI model training times.
Key Features of SK Telecom’s GPUaaS Platform:
• Scalable GPU Access: Flexible GPU allocation based on service scale and duration.
• Advanced Hardware: Currently powered by NVIDIA H100 GPUs, with H200 GPUs arriving in Q1 2025.
• Customizable Packages: Options include standalone servers, firewalls, and dedicated lines.
• Optimized Resource Management: AI Cloud Manager centralizes GPU performance management.
• Flexible Pricing: Pricing determined by contract length, number of GPUs, and prepaid billing models.
Kim Myong-gook, Head of SK Telecom’s GPUaaS business, stated: “The introduction of GPUaaS is a pivotal step in positioning SK Telecom as a major player in AI infrastructure. Our goal is to boost client competitiveness by offering a robust GPU platform for AI services.”
