LG CNS Unveils Pre-fabricated Modular Datacenter, Targets AI Factory Market
TECHWORLD ·
✦ AI Summary
LG CNS unveiled the Pre-fabricated Modular Datacenter (PMDC) and AI Factory full-stack service at Data Centre World Asia 2026 in Singapore.
PMDC is a method of pre-fabricating GPU, server, power and cooling equipment for on-site installation, allowing only the initially required scale to be built before modules are added as demand grows.
LG CNS has prepared a 4 MW module based on Vera Rubin and a 1.2 MW module based on GB300, and is pursuing expansion of its AI datacenter and AI factory businesses to Southeast Asia, North America and Europe.
LG CNS announced plans to expand its AI factory business. As expansion tools, it proposed a pre-fabricated modular datacenter and full-stack services based on NVIDIA's next-generation GPU.
On the 29th at Data Centre World Asia (DCWA) 2026 in Singapore, LG CNS unveiled the Pre-fabricated Modular Datacenter (PMDC) and its AI Factory full-stack service. To target the AI factory market, LG CNS presented both a datacenter deployment approach and its AI service capabilities.
PMDC is a method in which standardized modules for AI factory deployment, including GPU, servers, power equipment and cooling equipment, are pre-fabricated and then installed on site. The conventional approach first completes the datacenter building and then deploys mechanical, electrical and plumbing (MEP) systems and AI infrastructure. PMDC is designed to reduce on-site work compared with the conventional method.
This infrastructure is a phased expansion structure that can respond to changing demand, allowing only the initially required scale to be built first instead of making a large fixed investment from the outset. As demand for AI computing increases, modules can then be added to expand capacity, applying a modular addition approach based on the required computing scale.
LG CNS will pursue a PMDC-based 'modular AI factory' business by combining infrastructure capabilities across its group affiliates. In this process, LG CNS will handle datacenter design, deployment and operation (DBO), LG Electronics will connect modular cooling infrastructure, and LG Energy Solution will connect battery solutions.
The business is structured so that it can proceed without separately constructing a large-scale datacenter building. Instead, it has adopted a method of deploying container-type small datacenters.
LG CNS has prepared a 4 MW IT load module based on NVIDIA's next-generation GPU 'Vera Rubin' and also prepared a 1.2 MW module based on 'GB300.' Going forward, it plans to expand its lineup according to GPU configuration and power capacity.
LG CNS unveiled a Vera Rubin-based AI Factory full-stack service. Along with the unveiling, LG CNS presented the structure of its AI factory business.
The core component is the provision of GPUaaS. GPUaaS is a method of providing GPU resources via the cloud. The core package also includes an intelligence service that supports real-time training and inference for physical AI and other applications, integrated with power and cooling infrastructure.
LG CNS plans to expand beyond the model of separately providing GPU infrastructure and datacenter operations, moving into a business that integrates computing resources, AI services, and power and cooling equipment. It will also continue its business of converting existing datacenters into AI datacenters. The company said it plans to expand its AI datacenter and AI factory businesses through this exhibition, with target regions including Southeast Asia, North America and Europe.
Kim Tae-hoon, vice president and head of LG CNS's AI Cloud Business Division, said that an AI factory is a form that combines datacenters, GPU infrastructure and AI services. He added that the company aims to strengthen its competitiveness in related businesses based on its datacenter design, deployment and operation capabilities and its AI infrastructure expertise.
Source: TECHWORLD · Kim Seung-ki
Original: https://www.epnc.co.kr/news/articleView.html?idxno=407528
References
This article was produced with the help of an automated content generation algorithm.
Source: TECHWORLD
View originalThis article was summarized and organized by BizCrush based on the original article from TECHWORLD. For exact quotations and full details, please refer to the original article.