SK Telecom, Supermicro & Schneider Team Up

The collaboration leverages a modular, pre-fabricated model to construct AI data centers, integrating servers, power, and cooling into building-block style modules. This approach can slash deployment times by 30-60% compared to traditional on-site construction, which can take 18-24 months. The modular design also allows for phased deployment, enabling customers to scale their infrastructure as demand grows. Each partner brings a distinct expertise: SK Telecom provides its experience in AI data center operations, Supermicro supplies high-performance GPU servers optimized for AI workloads, and Schneider Electric delivers the mechanical, electrical, and plumbing infrastructure. This combination aims to address supply chain bottlenecks and enhance cost-efficiency for global hyperscalers and enterprises. The partnership directly targets the extreme power and cooling demands of AI workloads. AI-optimized server racks can require 40-100+ kW, a significant jump from the 5-15 kW used by traditional racks. Schneider Electric's contribution includes liquid cooling and power distribution solutions designed to handle high-density racks supporting over 1 megawatt per pod. The market for data center liquid cooling is expanding rapidly in response to these demands, with forecasts predicting it will grow from $5.52 billion in 2025 to over $18.79 billion by 2031. This growth is driven by the need to manage the heat generated by powerful AI accelerators and GPUs. SK Telecom is aggressively expanding its AI infrastructure, with plans to build over 1 gigawatt of hyperscale AI data center capacity in Korea to become a major hub in Asia. This includes a collaboration with OpenAI to build a new data center in the country's southwestern region. Supermicro specializes in high-performance servers for AI, offering systems optimized for large-scale training and inference with the latest NVIDIA GPUs and liquid cooling options. The company is also focusing on "sovereign AI," helping nations build their own AI infrastructure, and developing power-aware data centers that can interact with the energy grid. Schneider Electric's EcoStruxure platform provides the underlying infrastructure management, using IoT and analytics to optimize power, cooling, and overall operational efficiency. Their prefabricated modular systems are designed to be more energy-efficient, potentially achieving better Power Usage Effectiveness (PUE) ratios than traditional data centers. This collaboration reflects a broader industry trend of moving data center construction off-site to controlled factory environments. This method enhances quality control, reduces on-site labor needs, and allows for rigorous testing before modules are delivered, ensuring faster and more reliable deployment.