Collaboration as Part of Krafton's AI-First Strategy
Supported by Proven Experience in Reliable GPUaaS Services
Scheduled for Completion in July, Full-Scale Operation to Begin
NHN Cloud announced on January 19 that it has been selected as the operator for Krafton's large-scale Graphics Processing Unit (GPU) cluster project.
This project was launched to address the increasing demand for AI computing as Krafton began to expand its use of artificial intelligence (AI) following its declaration of an "AI-First" management strategy in October of last year. Krafton plans to invest approximately 100 billion won to build the GPU cluster as part of this strategy.
NHN Cloud has been chosen as the final operator for Krafton's GPU cluster project and has signed a contract with Krafton to provide infrastructure based on GPU as a Service (GPUaaS).
The GPU cluster is a core foundation for reliably executing Krafton's mid- to long-term AI strategies, including the advancement of AI platforms and the integration and automation of data. By leveraging NHN Cloud's GPUaaS, Krafton is expected to flexibly scale its infrastructure and enhance operational efficiency.
NHN Cloud independently designed and built the Gwangju National AI Center and has been providing stable GPUaaS services since 2021. Its years of experience operating AI infrastructure services for numerous public and private clients is cited as a key reason for being selected as a partner for this project.
NHN Cloud will establish a large-scale AI computing environment by configuring a GPU farm composed of over 1,000 units of NVIDIA's latest Blackwell Ultra GPUs in a multi-cluster structure. An XDR-800G ultra-high-speed InfiniBand network will be implemented to enable high-volume data transfers between GPUs with low latency and high speed.
This infrastructure will be built at the NHN Cloud Center (NCC) located in Pangyo, Gyeonggi Province. Pangyo NCC is considered well-suited for handling large-scale AI workloads, as it is equipped with power and cooling infrastructure for high-density computing and excellent network connectivity. The infrastructure is scheduled to be completed and become fully operational in July.
A Krafton representative stated, "The GPU cluster is a core foundation for company-wide AI operations as we pursue our AI-First strategy. We expect that NHN Cloud's outstanding GPUaaS services will strengthen the scalability and efficiency of our AI research and services overall."
An NHN Cloud representative commented, "This project demonstrates NHN Cloud's technological capabilities and operational expertise in building and running large-scale GPU clusters in the form of GPUaaS. We will focus on providing stable cluster operations and technical support to substantively back Krafton's AI-First strategy."
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
