NHN Cloud Recruits Users for National AI Data Center High-Performance Computing

Specialized Support for Research and Development of Ultra-Large AI Models

NHN Cloud announced on the 9th that it is recruiting users for the 'National Artificial Intelligence (AI) Data Center' HPC service until the 15th as part of the 'High Performance Computing (HPC) Service Regular User Recruitment' promoted by the Artificial Intelligence Industry Convergence Business Group (AICA).

Exterior of the National Artificial Intelligence Data Center
Photo by NHN Cloud

This project utilizes an accelerator composed of 304 NVIDIA H100 Graphics Processing Units (GPUs) with a performance level of 20 petaflops (PF) and a total storage capacity of 22 petabytes (PB). Based on this, it plans to specialize in supporting research and development of ultra-large AI models ranging from large language models (LLM) to small language models (sLM).

The recruitment targets domestic companies, public institutions, research institutes, universities, and associations that wish to research and develop products and services using ultra-large AI models. User selection is conducted by evaluating policy alignment, project excellence, resource necessity, target performance, and utilization plans. Selected companies and institutions will receive, free of charge for up to three months according to their submitted research and development plans, ▲large-capacity (2.5PF to a maximum of 10PF) or medium-to-small capacity (0.5PF to 2PF or less) accelerators and ▲storage space (30TB SSD and 200TB HDD per project).

Kim Dong-hoon, CEO of NHN Cloud, said, "In this project, NHN Cloud provides a high-performance HPC service optimized for strengthening AI model capabilities," adding, "NHN Cloud will continue to expand the AI ecosystem based on powerful AI infrastructure."

NHN Cloud has established and operates the 'National AI Data Center' in Gwangju Metropolitan City. The National AI Data Center is an AI-specialized data center optimized for AI research and development, housing the largest-scale HPC cluster in Korea. It is characterized as a high-spec AI data center equipped with computing power of 88.5PF and storage capacity of 107PB to support AI research and development for companies, research institutions, and universities. In addition, NHN Cloud is leading the AI infrastructure market by building a multi-AI GPU farm totaling 99.5PF, including more than 1,000 NVIDIA H100 GPUs at its own data centers including the National AI Data Center, 77.3PF based on NVIDIA, 11.2PF based on Graphcore, and 11PF based on Sapeon.