AI is a climate challenge. The solution? Nordic data centers

According to a new study by PriceWaterhouseCoopers, almost 50% of CEOs do not believe their businesses will be viable in a decade as tech and climate pressures accelerate.

The pressure to embrace new technologies such as Artificial Intelligence (AI) and continued digital transformation efforts to remain competitive, comes at a time when awareness of the worsening climate is also demanding action from business leaders.

The huge requirement of energy to fuel the rise in AI technology is in some cases, completely at odds with any ESG and related sustainability targets, not to mention the financial impact of sky-high energy costs.

Digitalization is not going to stop so it would seem that business leaders have an almost impossible task to address both these very different issues.

Yet there is a solution, and the answer lies with the data center.

Digital Infrastructure and the Impact on the Environment

Every digital transaction, from streaming online content to internet banking, starts a process in a data center and these processes necessitate extensive digital infrastructure that requires significant amounts of energy to power and cool. In fact, cooling is estimated to be responsible for 40% of data center energy usage.

This is not a new phenomenon. In 2019 researchers at the University of Massachusetts found that training several large common AI models can emit more than 283,948kgs of carbon dioxide – nearly 5 times the lifetime equivalent of the average car. Yet, up until recently it has been a problem only for very large data driven industries.

This is not the case anymore. As the use of AI becomes more widespread across all industries, the demand for energy to power the associated data centers will become truly colossal – and if those data center are coal fired, the impact of this trajectory is simply not sustainable.

Why It Matters Where Your Data Is

For many businesses the geographical location of their data may not be a top priority and yet it is an important factor in both sustainability and cost saving initiatives. Choosing a data center located in regions that benefit from a surplus of renewable energy sources such as hydro and geothermal energy can offer a stable long term power supply that is significantly cheaper compared to its natural gas counterparts.

Data center facilities in all of the Nordic countries benefit from a cool and more consistent climate that ensures that temperature and humidity levels within the data centers are maintained more efficiently, reducing energy outputs and ultimately decreasing carbon emissions. These conditions also enable the implementation of more energy efficient cooling technology systems such as Direct Liquid Cooling.

Nordic countries not only have abundant renewable and low-cost energy sources but they also have the infrastructure in place to enable circular economy initiatives. This means that modern data center campuses in these regions can be built with heat recovery technology from the outset. For example, atNorth’s SWE01 site in Stockholm, Sweden, partners with local energy provider, Stockhom Exergi allowing their district heating plant to draw the excess heat from the data center and recycle it to potentially heat up to 20,000 homes in the local community.

These countries also benefit from excellent connectivity, low latency networks and compliance with EU data sovereignty laws making them accessible and secure locations for data centers in addition to their environmental advantages.

This relatively simple change of data center location means that businesses can achieve best in class infrastructure to support technological advancements but also positively impact the bottom line - and go a long way to meet sustainability targets too.

By way of example, both BNP Paribas CIB and Shearwater Geoservices migrated portions of their IT workload to atNorth’s sites in Iceland, with BNP reducing energy consumption by 50% and decreasing CO2 emissions by 85% and Shearwater achieving a 92% reduction in CO2 output and an 85% reduction in cost.

‍