IDC: Moving to the cloud could save 1B+ metric tons of CO2 emissions

IDC has released a new study, the first of its kind, that predicts that cloud computing could help to prevent more than one billion metric ton of carbon dioxide (CO2) emissions in the next three years.

The forecast uses IDC data on server distribution and cloud and on-premises software use along with third-party information on data center power usage, carbon dioxide emissions per kilowatt-hour, and emission comparisons of cloud and non-cloud data centers.

IDC’s forecast includes upper and lower bounds for the estimated reduction in emissions. If the percentage of green cloud data centers today stays where it is, just the migration to cloud itself could save 693 million metric tons over the four-year time period.

If all data centers in use in 2024 were designed for sustainability, then 1.76 billion metric tons could be saved. IDC’s projection of more than 1 billion metric tons is based on the assumption that 60% of data centers will adopt the technology and processes that underlie more sustainable, “smarter” data centers.

“The idea of ‘green IT’ has been around now for years, but the direct impact of hyperscale computing can have on CO2 emissions is getting increased notice from customers, regulators, and investors, and it’s starting to factor into buying decisions,” said Cushing Anderson, program vice president at IDC in a statement.

“For some, going ‘carbon neutral’ will be achieved using carbon offsets, but designing datacenters from the ground up to be carbon neutral will be the real measure of contribution. And for advanced cloud providers, matching workloads with renewable energy availability will further accelerate their sustainability goals.”

The reduction in emissions is largely predicated on the theory that aggregated compute resources across enterprise and large data centers, especially cloud data centers, are more efficient. That efficiency can result in more efficient power capacity, cooling, and server utilization.

It’s not that far out of a theory. Hyperscalers like AWS and Google tend to cycle through hardware a lot faster than enterprises and especially SMBs. It’s not unusual for a hyperscale cloud provider to retire servers after one or two years and sell them used, while businesses tend hold on to servers for five years or more.

IDC says that the biggest opportunity lies in those areas that generate higher emissions per kilowatt-hour. American hyperscalers are exceptionally green-minded and usually put their data centers close to hydro- and wind-energy resources.

Not so much in the Asia-Pacific. Most of their data centers are powered by coal-fired plants. IDC said if APAC nations move off coal, it could save more than half of the total emissions over the next four years. On the other hand, Europe, Middle East and Africa (EMEA) have low emission per kilowatt-hour, which means it will generate just 10% of total savings.

IDC also said that while shifting to cleaner sources of energy is very important to lowering emissions, reducing wasted energy will also play a critical role. Cloud data centers are doing this through reducing the energy spent cooling them and optimizing the physical environment. “The goal of an efficient data center is to have more energy spent on running the IT equipment than cooling the environment where the equipment resides,” the report stated.

And that is one thing hyperscalers excel at because every dollar saved through energy efficiency goes to the bottom line. Energy efficiency isn’t as top of mind for private data centers.

Another advantage for cloud-computing providers is the ability to shift workloads to any location around the globe, another trick to lower CO2 emissions. Heavy workloads can be moved–within compliance limits on data–to more efficient locations.

The IDC report, Worldwide CO2 Emissions Savings from Cloud Computing Forecast, 2021–2024: A First-of-Its-Kind Projection (IDC #US47426420), provides annual CO2 savings on a worldwide and regional basis and estimates the annual savings for each year above the 2020 baseline.