SK Telecom, South Korea’s largest wireless carrier, has announced that it will be opening a new artificial intelligence (AI) data center in Seoul in December. The center will be built in partnership with Lambda, a US-based graphics processing units (GPUs)-as-a-service provider.
The partnership aims to capitalize on the growing demand for AI services in South Korea. By establishing a large-scale Nvidia GPU cluster, SK Telecom and Lambda will provide businesses with the necessary computing power to develop and deploy AI applications.
Lambda’s AI cloud services utilize the latest GPUs from Nvidia, offering high-performance computing capabilities. The company boasts a clientele that includes major tech giants like Intel and Microsoft.
The new AI data center will be located in the Gasan data center of SK Broadband, SK Telecom’s internet service provider. The facility will feature Nvidia GPU clusters running on Lambda’s cloud computing platform. SK Telecom plans to significantly expand the number of GPUs in the center over the next three years, keeping pace with the increasing demand for AI resources.
To ensure the stable operation of the GPU servers, SK Broadband has implemented a power density of 44 kilowatts per rack, which is approximately nine times the average power density in domestic data centers.
With the launch of the AI data center, Lambda will be expanding its presence to the Asia-Pacific region for the first time. Domestic companies will now have the option to store their data locally using Lambda’s GPU-based AI cloud services.
SK Telecom also intends to introduce a GPU-as-a-service (GPUaaS) offering, providing businesses with access to GPU resources for computation and rendering purposes. This service will leverage Lambda’s GPU infrastructure.