Register for free and continue reading
Join our growing army of changemakers and get unlimited access to our premium content
Could this platform cut the carbon footprint of energy-intensive computing processes?
Spotted: As generative AI becomes mainstream, there is growing urgency to build out the Cloud infrastructure needed to support the compute-intensive large language models used in AI.
In particular, there is a need for more energy-efficient data centres — the International Energy Agency projects that by 2026, global data centres will use 1,000 terawatts of electricity, roughly equivalent to the total power consumption of Japan.
Cloud provider Crusoe Energy Systems is taking an innovative approach to delivering sustainable computing infrastructure. Crusoe points out that a lot of energy is currently wasted or ‘stranded’. This includes natural gas flaring, or wind turbines and solar farms that need to be shut down when they produce more power than the local grid can cope with.
The company plans to co-locate modular data centres near these sites in order to take advantage of the energy that would otherwise go to waste. This not only allows Crusoe to offer Cloud computing with a lower carbon footprint than traditional infrastructure, but can also help to stabilise electrical grids by using excess power. Its cloud platform, CrusoeCloud, is designed to support the most cutting-edge and energy-intensive applications, including AI workloads, graphical rendering, computational biology, therapeutic drug discovery, and simulations.
In July this year, Crusoe announced that it would be building a 200-megawatt data centre at energy tech company Lancium’s Clean Campus in Texas. The centre will allow Lancium to run heavy AI workloads using renewable energy and is expected to be up and running in 2025.
Written By: Lisa Magloff