Data centres are the beating heart of the digital age. They store, process, manage, and disseminate the huge amounts of digital data that fuel our daily lives, from online searches and social media to cloud services and, more recently, Generative Artificial Intelligence (GAI) tools such as ChatGPT and other AI. However, this gigantic computing power comes at significant environmental cost, particularly in terms of water consumption. Although there is enough freshwater on our planet to support the world’s population of around 7 billion people, this resource is distributed unevenly, and much of it is wasted, polluted, and unsustainably managed. Some of the world’s leading technology companies have been looking for innovative ways to reduce their water footprint. Could quantum computing provide a promising solution in the coming years?
‘Thirsty Tech’ - the excessive water consumption of data centres
Data centres are notorious energy guzzlers, necessitating huge amounts of electricity to feed their servers and maintain optimum temperatures. The data centre industry today accounts for around 4% of global electricity consumption. However, a lesser-known aspect of their environmental impact is their demand for water. In a similar way to a nuclear power station, the heat generated by data centres means that they need to be cooled constantly, and water is often used to achieve this. Water is circulated to cool the servers and maintain the infrastructure at acceptable temperatures, but this process has a significant impact on local water resources.
Google and Microsoft, to name but a few, operate some of the largest data centres in the world. They are spread over different regions, each with its own set of challenges in terms of water consumption, making water management a major focus for these technology companies.
The success of Generative AI is amplifying the problem
Generative Artificial Intelligence and data centres have considerably boosted the development of the digital industry in recent years, but this rapid growth has also had a significant impact on the environment.
In response to questions from the Associated Press, Associate Professor Shaolei Ren, an expert in IT at the University of California, and his team have calculated that for each series of 5 to 50 questions submitted to ChatGPT about half a litre of water is consumed. This figure may vary depending on the location of the servers and the time of year, but the calculation also takes into account the indirect consumption of water that companies do not measure, for example water used to cool the power stations that supply electricity to the data centres.
This finding is corroborated by the internet giants themselves: in its most recent environmental report, Microsoft revealed that between 2021 and 2022, its global water consumption had increased by 34%. Over the same period, Google reported a 20% increase in its water consumption. According to these two industry giants, these significant increases can largely be attributed to their work on AI.
The ‘Water Positive’ intentions of the internet giants
The Water Positive intentions for data centres aim to promote the responsible and sustainable use of this valuable resource within IT-critical infrastructures. This includes the deployment of more efficient cooling technology, such as the reuse of cooling water, the adoption of atmospheric cooling techniques, and the use of recycled water for cooling operations. Furthermore, some data centres are investing in clean energy sources to power their seawater desalination plants. They are even collecting rainwater and closely monitoring their water consumption to identify any leaks. By working with water conservation organisations and incorporating sustainable practices into the design of new facilities, data centres are striving to minimise their environmental impact while continuing to provide essential data storage and processing services. Google aims to achieve a target of replenishing 120 per cent of the freshwater volume it consumes by 2030. In 2020, Microsoft also announced its undertaking to achieve a positive net water footprint by 2030 and joined the Water Resilience Coalition. Amazon Web Services (AWS), for its part, has joined the non-profit organisation Water.org, co-founded by Hollywood actor Matt Damon, driven by the shared belief that solving the global water crisis is possible.
Quantum Computing: A glimmer of hope for 2030
This revolutionary technological advance offers a glimmer of hope for the reduction of water consumption in data centres. Unlike traditional computers, quantum computers use qubits that draw on the principles of quantum mechanics. This means that complex calculations can be carried out much faster than with traditional computers.
Quantum technology could help to reduce water consumption in a number of ways. Quantum computers may be capable of solving complex problems relating to the optimisation of heat management in data centres. This would reduce reliance on water-cooled air conditioning, saving significant amounts of water. They should also be able to speed up data analysis, meaning that complex data processing tasks can be completed more quickly, in a matter of hours instead of years, thereby reducing the time servers run and consequently their demand for energy and cooling.
Although these quantum computers themselves have demanding cooling requirements and must be maintained at temperatures colder than interstellar space, early research in this area suggests that the overall energy consumption of quantum computers will be three orders of magnitude lower than that of conventional machines.
Towards less water-thirsty and more sustainable data centres
Water consumption in data centres has become a major environmental issue, with technology giants such as Google and Microsoft looking for solutions to reduce their water footprint. The rise of Generative Artificial Intelligence (GAI) has amplified this problem with its increased demand for energy and water. However, the efforts being made and the advent of quantum computing offer cause for hope. This revolutionary technology could significantly improve the energy efficiency of data centres by solving complex heat management problems and speeding up data processing, thereby reducing their need for water-cooled air conditioning. Although quantum computers require customised cooling, their overall energy consumption is expected to be significantly lower than that of traditional computers. By combining these advances with positive water-related initiatives, data centres can look forward to becoming more sustainable and environmentally friendly by 2030, helping to meet the challenge of ‘Thirsty Tech’.
