NHN Cloud Gwangju National AI Data Center / Courtesy of NHN Cloud

NHN Cloud stated on the 1st that it is leading the artificial intelligence (AI) infrastructure business based on the operation of the national AI data center.

NHN Cloud, which operates the Gwangju National AI Data Center, is a prominent cloud service provider (CSP) that currently offers powerful GPU as a service (GPUaaS) through the national AI data center, which is capable of providing the GPUs necessary for AI development and research in a cloud format.

The national AI data center operated by NHN Cloud provides large-scale AI specialized infrastructure services that can support AI research and development and service operations for corporations, research institutions, and universities. The computing power is 88.5 petaflops (PF), and the storage capacity is 107 petabytes (PB). An 88.5 PF computing power is capable of processing the equivalent of about 500,000 general-purpose laptops in one second, while the 107 PB storage capacity corresponds to the storage capacity of 107,000 one terabyte (TB) hard disks.

NHN Cloud has garnered attention in the industry by securing a large quantity of NVIDIA's high-performance GPU 'H100' as the first in Asia amid the heated competition for AI semiconductor acquisition due to the global AI development boom. Along with this, it has a well-rounded collection of domestically developed neural processing units (NPUs) and various types of high-performance GPUs, providing an optimized infrastructure for AI research and development that supports deep learning training, data analysis, and utilization of vast amounts of data in a short period.

In particular, NHN Cloud has exceptional GPU operational capabilities through large-scale actual operations in the domestic market. High-performance GPUs like the H100 require a high level of power to operate efficiently, and the national AI data center has created an environment that supplies 15 kilowatts of power per rack that can handle this demand. Additionally, it has established a cooling system capable of efficiently dissipating the high heat output of GPUs by adopting technologies such as adiabatic cooling systems and using high-capacity fans and cooling coils to supply cool air without interruption and maintain high performance.

A representative from NHN Cloud noted, "We have proven to the market that we are a capable operator of large-scale GPU acquisition and operation and have accumulated sustainable GPU operational know-how and performance optimization experience. Based on this infrastructure and know-how, we are supplying GPUs to institutions such as the Meteorological Science Institute and NCSOFT, and we plan to participate in numerous GPU supply projects led by the government, local governments, and public institutions to lead the revitalization of the domestic AI ecosystem and enhance national AI competitiveness."

※ This article has been translated by AI. Share your feedback here.