As the global artificial intelligence (AI) bubble controversy intensifies, Ellice Group is gaining attention for its expansion into modular data centers, driven by robust AI demand. The company aims to become a global AI cloud enterprise by targeting both domestic and international markets.
Kim Jae-won, CEO of Ellice Group, stated on November 11, "There is no disagreement that AI infrastructure is necessary. While companies respond to the market in different ways, we are focused on providing stable infrastructure that meets demand."
On the first day of its KOSDAQ listing on October 20, Ellice Group's shares closed at 61,500 won, down 32.04% from the initial public offering price of 90,500 won. This decline was influenced by a sharp drop in the bond value of U.S. cloud company CoreWeave and heightened concerns over funding burdens associated with large-scale AI infrastructure investments.
Kim addressed the volatility in stock prices at the time of listing, saying, "The AI infrastructure market is highly volatile due to its significant growth potential. There were concerns in the global market at the time of our listing."
He explained, "While CoreWeave aggressively utilized financing to increase its market share, we focus on providing infrastructure based on demand forecasts from a modular perspective. By building incrementally like stacking Lego blocks, we can reduce excessive financial burdens and liquidity risks."
Ellice Group possesses full-stack capabilities in AI infrastructure. The company began designing modular data centers in 2021 and completed its first construction in 2022, accumulating five years of experience in GPU infrastructure operations. Its structure integrates everything from data center hardware to infrastructure management software (ECI), GPU cloud, and industry-specific AI transformation (AX) solutions.
Kim noted, "Creating containers or developing cloud software is different from integrating them into a single full-stack to provide stable services. Our technological expertise, built through five years of operational experience and trial and error, is our competitive advantage."
Currently, GPU cloud utilization has reached 100%. With supply not keeping pace with demand, the company plans to expand its infrastructure based on funds secured since its listing, aiming to maintain an 80-90% utilization rate while scaling up. The growth of its cloud business, which accounts for about 80% of revenue, will also enhance synergies with the AX business.
Ellice Group is also intensifying its efforts to penetrate international markets. It plans to establish modular data centers in the U.S. and Singapore while allowing overseas customers to remotely access its cloud services. Expanding the proportion of private clients, which currently stands at 20-30%, is another challenge, as public projects account for 60-70% of its revenue structure.
Kim believes that the demand for AI infrastructure is still in its early stages. As AI applications expand in sectors such as healthcare, finance, and advertising, and as physical AI develops in manufacturing and construction, the demand for GPUs is expected to continue rising.
He stated, "AI is becoming an essential cost for companies to operate. Beyond semiconductor manufacturing competitiveness, global companies must emerge in cloud software as well. Ellice Group aims to grow into a global B2B software and cloud enterprise."
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.