Kakao Expands Energy Efficiency in Data Centers, Reducing Cooling Costs by 23%

By Shin Hye An Posted : September 22, 2026, 09:32 Updated : September 22, 2026, 09:32


As the demand for effective heat management and energy efficiency in data centers grows, Kakao is expanding its energy management capabilities from its operations to leased data centers.

On September 22, Kakao announced that it has improved the cooling system operations at its leased data center in Hanam, Gyeonggi Province, while maintaining service stability and reducing energy consumption for cooling.

This initiative applies the energy management experience gained from its own data center to enhance the operational efficiency of external leased facilities.

Kakao is committed to reducing the environmental impact of its data centers through ongoing efforts to improve energy efficiency as part of its ESG (Environmental, Social, and Governance) management strategy. At the Kakao data center in Ansan, various energy-saving technologies have been implemented, including cooling system optimization, air conditioning improvements, power equipment efficiency enhancements, and solar energy self-consumption, all aimed at ensuring stable server operations while reducing carbon emissions.

The company reported a total reduction of 4,753 tCO₂eq (tons of carbon dioxide equivalent) in greenhouse gas emissions over the past year. This figure is comparable to the annual emissions of approximately 3,400 passenger vehicles based on the average emissions of commercial vehicles as reported by the Ministry of Environment.

Kakao has applied its energy management experience from the Ansan data center to the leased facility in Hanam. Instead of replacing or expanding equipment, the company optimized the cooling temperatures of the server room and the chilled water temperature of the cooling units based on actual operational data and site conditions.

As a result, the facility maintained stable operations even amid record heat and fluctuations in server usage. Compared to the same period last year, cooling energy costs were reduced by about 23%, even when considering the impact of server usage and external conditions.

Kakao views this achievement as a significant collaboration that combines its data-driven heat management expertise with LG CNS's operational experience in data centers to reduce cooling energy consumption. Kakao proposed energy-saving measures, while LG CNS adjusted the operation methods of key cooling equipment, such as the fan coil units (FWU) and chillers, based on its operational experience and site knowledge, validating operational stability in stages.

Some server rooms have also implemented AI agents developed by LG CNS. These AI agents analyze the server room environment and operational data in real-time, autonomously controlling the cooling systems to maintain stable operations while reducing energy usage.

A representative from LG CNS stated, "This project exemplifies the collaboration between Kakao and LG CNS, prioritizing operational stability in data center management while applying existing equipment efficiency improvements and AI-based autonomous cooling technology. We will continue to support our customers in enhancing energy efficiency based on our expertise in data center operations and AI technology."

A Kakao representative added, "Having confirmed that energy efficiency can be improved based on operational data even in environments with limited equipment replacement, we will continue to expand eco-friendly data center operations based on Kakao's technological capabilities and on-site collaboration."

Meanwhile, Kakao is also increasing its procurement of renewable energy to reduce the environmental impact of its data center power usage. In 2025, Kakao introduced a solar power purchase agreement (PPA) for the first time at its Ansan data center, and in the same year, the total amount of renewable energy procured by Kakao across the company reached 29,903 MWh, an increase of approximately 2.8 times compared to the previous year.





* This article has been translated by AI.

Copyright ⓒ Aju Press All rights reserved.