The energy and emissions cost of digital transformation

We are going through an incredible digital transformation of the economy. Annual smartphone shipments have more than doubled since 2010, hitting 1.2 billion in 2023. Business e-commerce sales grew nearly 60% from 2016 to 2022, based on data from 43 countries representing about three-quarters of global GDP.¹ And since 2010, global internet traffic has expanded 25-fold.²

With the strong growth in demand for data centre services, energy demand for data centres and data infrastructure has grown as well, but not as fast. This is thanks in part to efficiency improvements in IT hardware and cooling, and a shift away from small, inefficient enterprise data centres towards more efficient cloud and hyperscale data centres. Even so, data centre energy use grew about 20% to 40% annually over the past several years.³

GHG emissions from data centres and infrastructure have grown only moderately, reaching about 1.4% of global GHG emissions⁴, as the operators of data centres and infrastructure invested heavily in renewable electricity. In fact, the world’s largest providers of cloud services (Amazon, Google, Microsoft) are also the largest corporate buyers of renewable electricity.  

The great AI acceleration

Now we are entering a new phase. With the rapid rise in AI, technology companies are no longer confident that their decarbonisation efforts will keep pace with the increase of computing power and associated energy demand.  

Early July, Google reported that its GHG emissions in 2023 had risen 13% compared with the previous year, hitting 14.3m metric tons⁵, with its investments in AI and associated additional data centre capacity as an important factor. It states that the uncertainty about the future environmental impact of AI made it more difficult to reach its highly ambitious goal of net emission zero by 2030. Only a week before, Microsoft also said that its ‘moonshot’ goal of being carbon-negative by 2030 may not succeed because its heavy investment in data centres and infrastructure for AI.⁶ 

The International Energy Agency projects that total electricity consumption from datacentres could double from 2022 levels to 1,000 TWh in 2026, equivalent to the energy demand of Japan.⁷

How companies can keep digital technology-related GHG emissions in check

This is not just a problem for the tech giants; it also affects the companies that rely on cloud solutions to run their operations, as the emissions are part of their scope 3 GHG emissions footprint.  

What can these companies do? Digital technologies are so fundamental to business that scaling back is not an option. But companies can make more conscious choices about their use of cloud and AI by integrating emissions considerations. 

1. Selection of your cloud services provider

Many companies have already moved workloads from on-sites servers to the cloud, but where this has not been done yet, it can drive down energy use and emissions significantly. Research by AWS suggests that hyperscale data centres can be infrastructure is up to 4.1 times more efficient than on-premises.⁸

Some cloud services providers have invested much more in energy efficiency and renewable energy than others, so their emissions per CPU are much lower. By taking the energy use and emissions intensity into account when choosing a cloud provider, companies can significantly reduce the GHG emissions footprint of their technology use.  

2. Software selection 

While measures to reduce energy consumption of data processing have focused on hardware solution, programming choices in the development of the software can also make a real difference. Programmers are looking to improve energy efficiency of software by improving data-collection and processing techniques, choosing more-efficient libraries, and improving the efficiency of training algorithms. Companies should therefore also consider energy use in their software selection.  

3. Applying AI to enable emissions reduction elsewhere

Additionally, companies should seek opportunities to apply AI for emissions reduction. Opportunities are endless, from energy efficiency in buildings, process optimisation, predictive maintenance, efficient routing of transport, precision agriculture, and many more.   

Technology can make it easier or more difficult to operate sustainably. The choice is up to us.

Author: Sytze Dijsktra, Netherlands Country Manager, Simply Sustainable

For more information and advice on your business’ digital strategy, please get in touch with one of our friendly experts. 

1. https://unctad.org/publication/digital-economy-report-2024

2. https://www.iea.org/energy-system/buildings/data-centres-and-data-transmission-networks

3. https://www.iea.org/energy-system/buildings/data-centres-and-data-transmission-networks

4. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4424264

5. https://www.gstatic.com/gumdrop/sustainability/google-2024-environmental-report.pdf

6. https://www.theguardian.com/business/article/2024/jun/29/ai-drive-brings-microsofts-green-moonshot-down-to-earth-in-west-london

7. https://iea.blob.core.windows.net/assets/18f3ed24-4b26-4c83-a3d2-8a1be51c8cc8/Electricity2024-Analysisandforecastto2026.pdf

8. https://www.aboutamazon.com/news/aws/aws-carbon-footprint-ai-workload