The development of the internet and cloud computing has significantly increased the amount of accessible information stored in data centers. Data centers consume a lot of energy, but they also help the economy save a lot of energy.
By facilitating digitalization across infrastructure and industry in general, this is accomplished. Through energy-efficient design, major IT corporations are accepting responsibility for the energy consequences of data center energy consumption. The average data center, however, is already more than 20 years old, and many of them were not built efficiently. The only way to keep costs down and reliability high is to continue to be at the forefront of energy efficiency. Numerous chances exist for system enhancement and huge energy savings.
What Are Data Centre Energy Consumption?
All data center energy consumption points are merely structures that house servers in variable numbers and offer room, electricity, and cooling for network equipment. The number of servers in a data center directly affects how much power it uses. Data centers centralize essential IT activities and resources while storing, managing, and sharing data. They can be grouped according to their location, ownership structure, and business model. Edge data centers, which are strategically positioned close to where the data is being collected and analyzed, are anticipated to experience the most significant increase in the data center industry and are projected to nearly quadruple by 2024. By 2025, outside of the typical data center or cloud, more than 75% of data generated by businesses will be created and processed.
How is Data Energy Measured?
There are presently no official data on data center energy consumption points accessible globally or nationally. As a result, mathematical models must be utilized to calculate the quantity of energy used. So-called “bottom-up” models consider the installed inventories of IT devices in different data centers and their energy consumption characteristics to estimate overall energy use. Bottom-up research offers a plethora of knowledge regarding the factors influencing energy usage, but it is uncommon because it takes a lot of effort and data. According to the most reliable bottom-up analysis of the past ten years, data centers consumed between 1.1 and 1.5% of the world’s electricity in 2010.
Contrarily, extrapolation-based models forecast overall energy consumption by scaling up earlier bottom-up projections based on data center market growth indicators like global IP traffic or data center investments. Since extrapolation-based processes are easier to use, they have been used to fill in the gaps left by sporadic bottom-up research. Such extrapolations frequently predict significant increases in data center energy consumption demand since the market indicators are expanding rapidly upon which they are based. By extrapolating from past statistics, several often-cited extrapolations have suggested that the energy used by data centers worldwide may have doubled since 2010 and would likely continue to climb quickly in the coming years. These statistics have received a lot of media attention, confirming the widely held belief that data center energy use is rising more quickly than the demand for data.
Despite the sharp rise in demand for information services in recent years, new bottom-up data indicates that the energy consumption of data centers worldwide only climbed by 6% between 2010 and 2018. These new conclusions differ from the earlier study because they consider several recent statistics that more accurately describe the installed stocks, operational traits, energy consumption of data center IT equipment, and structural changes in the data center industry.
The conclusion that worldwide data center energy consumption probably consumed about 205 terawatt-hours (TWh), or 1% of global power use, in 2018 is significantly different from past extrapolation-based estimates that suggested rapidly growing data center energy use over the last decade. The near plateau in energy use attributes to three key efficiency effects:
- Servers and storage drives are now much more energy-efficient because of continual technological advancements by IT manufacturers.
- The energy consumption of each hosted application has significantly decreased due to the increased use of server virtualization software, which enables multiple programs to run on a single server.
- Most compute instances have been transferred to the massive cloud and hyper-scale data centers, which use energy-saving ultra-efficient cooling systems among other things.
What is the Way Forward?
The development of more reliable and predictive methods that increase the frequency of bottom-up insights and do away with the drawbacks of extrapolation-based forecasts is a key priority for the energy analysis community as decision-makers need to evaluate future efficiency and mitigation options confidently.
