Data centers have come under fire for their climate footprint for decades, and yet they still make up a large share of global IT emissions. New research has uncovered an unexpected reality. Carbon emissions increasingly, the true culprit behind rising carbon emissions is end-user devices. Public awareness of sustainability issues is rapidly increasing. Given the current state, it is more important than ever that we understand the impact all IT emissions have, including those from manufacturing and device operation.
Based on their latest research, data centers make up around 40 percent of all IT emissions. By comparison, end-user devices like smartphones and laptops account for a shocking 60 percent. That means that devices worldwide are responsible for 1.5 to 2 times the carbon emissions of all data centers together. These data center operations only contribute to 24 percent of total emissions. Far more damaging, we can no longer ignore the environmental impact created by personal devices.
The Emission Breakdown
Now, a deeper dive into the data reveals that data centers have been in the limelight for their energy use. In truth, the manufacturing process for devices is the much greater threat to our environment. With about three-quarters of a device’s total emissions coming from production, this makes the case for cleaner manufacturing more urgent than ever.
For example, the production of one smartphone creates an impact of around 50 kg CO2e. In the case of laptops, this number jumps to 200 kg CO2e. These electronics emissions are an eye-opening reminder of the harmful impact that our reliance on electronic devices has on the environment. This combination pressures producers into adopting more sustainable practices.
Moreover, when considering embodied carbon—the total carbon emissions associated with the entire lifecycle of a product—devices reveal an even greater environmental impact. The embodied carbon footprint of data centers is 16 percent. Devices make up a much greater and alarming 45 percent. This troubling juxtaposition illustrates that the environmental burden of these devices extends past merely operating these devices. It’s in large part because of how they manufacture their trucks.
The Potential for Change
The trend of device discussions online and in advocacy circles gets more serious when considering plans to reduce total carbon footprints. One major opportunity is in lengthening the lifecycles of smartphones. According to our research, increasing the average smartphone lifecycle from two years to three would cut annual manufacturing emissions by one third.
With over 1 billion smartphones replaced every year, there is plenty of potential. This replacement would result in a beautiful reduction of more than 50 million tonnes of CO2e per year, just from the production of cellphones. This kind of shift is not just good for the environment – it can save consumers money.
Beyond lifecycle extensions, energy efficiency has to become top of mind for these companies. Even further, Google just a while back announced that their datacenters are 1.5 times more energy-efficient than the industry average. Organizations working towards digital sustainability can no longer solely focus on improving data center efficiency, but rather need to combine these efforts with device-related emissions reductions.
Regulatory Landscape and Future Outlook
As businesses continue to grapple with the challenges of sustainability reporting, the regulatory landscape is shifting to demand accountability for emissions. The new Corporate Sustainability Reporting Directive (CSRD) expanded EU requirements to disclose emissions to nearly 11,700 companies. Unfortunately, most of them rely on frameworks that are not truly complete. This regulatory change opens the door for institutions to take a more holistic approach in understanding and managing their impacts on the environment.
For a glimpse into the new complexities around energy consumption, artificial intelligence (AI) workloads are another piece of the puzzle. New projections indicate that AI workloads alone may account for as much as 12 percent of U.S. electricity usage by 2028. With AI’s rapidly growing footprint across the economy and society, accounting for and reducing its environmental impact will be increasingly important.