Fujitsu Develops Highly Accurate Predictive Technology to Reduce Energy Consumption in Data Centers

KAWASAKI, JAPAN: Fujitsu Laboratories announces development of highly accurate predictive technology for metrics such as temperature and humidity that reduces the electricity used by air conditioning equipment in data centers by 20 percent.

With the growing demand for data centers increasing, the energy consumed by them also increases. Data centers account for 1-2 percent of all electricity used, out of which 30-50 percent of energy consumed is by air conditioners used in data centers. 

Existing data center air conditioners function on the basis of sensors installed to maintain the target temperature and humidity when operating temperatures of equipment such as server increases. When sensor data exceeds a set value, the system carries out rapid cooling for safety reasons, which causesair-conditioning to operate excessively. This increases the energy consumed.

Fujitsu has developed a highly accurate predictive technology that can respond to the frequent change in the data center by building sequential models. Fujitsu applied Just in Time (JIT) modeling, which can build a model in response to a situation.

Through existing JIT modeling, useful information including temperature and humidity is selected from sensor data through mathematical processing, and this data is used to create a predictive model and calculate predicted values. However by using data from sensors only, the method does not select enough useful information to immediately reflect the status of dynamic changes in the data center limiting highly accurate predictions. Fujitsu’s new technology creates a database that incorporates the status of air-conditioning equipment, utilization rate of machines and fan speeds of the air conditioner, then selects useful information against prediction targets and automatically choose at least one variable from the air-conditioning equipment status database thus saving energy consumption by up to 20 percent.

This new technology can predict the temperature of servers' air supply with high accuracy, within an average margin of ± 0.17°C and maximum ± 2.1°C, even in frequently changing environment conditions. Fujitsu with this technology aims to save energy in data centers and reduce the effects of global warming on the globe and environment.