NTT Data and Daikin will begin joint verification of a next-generation data center cooling optimization solution in July 2026. The companies will develop AI that predicts thermal conditions inside servers and use the results to coordinate air-conditioning, heat-source and liquid-cooling equipment.
The AI will estimate internal server thermal conditions from indirect data, including server power consumption and temperature information, without directly obtaining detailed internal server data. It will learn the relationship between server load and thermal conditions to predict conditions closer to those inside the server.
Based on these predictions, the system will integrate control of air-conditioning, heat-source and liquid-cooling equipment. The companies aim to reduce excessive cooling, improve response to load fluctuations and optimize energy use across the data center rather than individual systems.
NTT Data will provide data center operating expertise, correlation data covering server behavior and air conditioning, its GDCEMS energy visualization and analysis system, and the verification environment. Daikin will provide equipment control technologies, an autonomous AI-based air-conditioning control device, and support equipment operation and performance verification. Both companies will jointly develop the thermal prediction AI.
The solution will be applied at an NTT Data data center from July 2026 to March 2027. The verification will assess energy savings, electricity cost reductions, operational automation and the accuracy of AI-based thermal predictions. The companies aim to commercialize the solution during fiscal 2027.
