This post is also available in:
עברית (Hebrew)
A European initiative is studying the practicality of building data centers in space, and it has recently found that the project could be economically viable.
The reason for trying to build data centers in space is that they could significantly help reduce the carbon footprint of the infrastructure that is powering the AI intelligence boom. Aerospace company Thales Alenia Space (the coordinator of the project) said in a statement that a Horizon Europe-funded feasibility study confirmed deploying them into space could offer a more sustainable solution for hosting and processing data, as well as generate a return on investment of several billion between now and 2050.
These space centers could also have a significant environmental impact – while data centers provide the computing power that AI models need to run, we also know that they consume enormous amounts of energy and water. And so, the recent rise of AI applications signaled the beginning of a race among tech companies to build more data centers, which in turn raised concerns about the potential environmental impact.
According to Innovation News Network, this study set out to compare the environmental impacts of space-based and Earth-based data centers, and moving forward, the company plans to consolidate and optimize its results.
Experts suggest that space data centers could be powered by solar energy outside the Earth’s atmosphere and contribute to the EU’s goal of achieving carbon neutrality by 2050. Thales Alenia CTO, Christophe Valorge, stated: “The results of the Ascend study confirm that deploying these centers in space could transform the European digital landscape, offering a more eco-friendly and sovereign solution for hosting and processing data.”
However, the study found that space data centers would require developing a launcher that is ten times less emissive over its entire lifecycle in order to significantly reduce the carbon dioxide emissions that are generated by the processing and storage of digital data.
