Huawei is helping build green data centers. Artificial intelligence2021-01-04
Data centers are being overhauled, renovated and modernized to support digital transformation and their energy-consuming cloud deployments.
Artificial intelligence has become an important part of this revision and the most important issue for data centers by addressing both cost and environmental issues at the point of energy saving. Artificial Intelligence solutions are used to make data centers more energy efficient, safer and more reliable.
Huawei used machine learning to develop the iCooling intelligent thermal management solution for data center infrastructure and solve the energy consumption problem. The iCooling system uses deep learning to analyze historical data and identify key factors affecting energy consumption and generate PUE prediction. An optimization algorithm then determines the ideal parameters that are transmitted to various control systems.
At Huawei’s cloud Data Center Langfang in North China, the use of iCooling resulted in an eight percent lower PUE than before, which translates into a huge savings in annual power costs. And in a Chinese mobile data center in Ningxia, the use of iCooling technology reduced the data center’s total energy consumption by 3.2 percent and saved more than 400,000 kWh of electricity per year. According to Huawei, as data center loads increase and Artificial Intelligence learning capability improves, six million kWh of electricity will be saved in the data center each year, equivalent to about three million kilograms of carbon dioxide emissions.
While advanced power management is a striking example of the value of AI for the data center industry, AI is proving extremely valuable in other ways in managing data centers. It is used to increase the security and reliability of data centers.
Prior to the introduction of AI, fault detection was addressed on a case-by-case basis, isolating the defective component. By gathering information from power supply and distribution systems, Artificial Intelligence can predict impending device and component failure, alert operations and maintenance personnel, and provide additional information to aid decision making.
For example, Huawei’s iPower intelligent power supply and distribution technology can detect and isolate faults and then restore them at sub-second speeds, improving the reliability of the data center’s power infrastructure and reducing the risk of fire. iPower also accurately predicts battery life and health and provides preventive maintenance before any malfunctions occur.
Artificial intelligence can also be used to reduce repetitive work in the data center and to monitor and manage fragmented resources that can be easily ignored by other methods. Huawei’s iManager data center infrastructure management system uses smart hardware and IoT sensors to manage energy, cooling, and space to optimize utilization, and Artificial Intelligence manages the distribution and operation of assets. It is estimated that iManager increased the resource utilization rate by 20 percent. iManager also supports centralized management and network management for multiple data centers in different locations. As edge data centers grow in popularity, iManager can be used to reduce the need for field visits and thus increase the efficiency of data center management.
Research by the Mckinsey Global Institute shows that artificial intelligence has the potential to contribute approximately USD 13 trillion to global economic activities by 2030 and increase global GDP by 1.2% per year. Artificial intelligence is driving changes in the world with profound and potentially big impact, and data centers are no exception. The use of artificial intelligence provides the opportunity to establish safer, more reliable, energy efficient and cleaner data centers. According to M Capital Group, the data center industry is projected to grow by up to 15 percent in the next five years, which is a very good development.