09 October 2024
Andy Connor, Channel Director EMEA, Subzero Engineering
Over the last few years and looking to the near future, the demand for data processing has soared exponentially at rates previously thought inconceivable. Goldman Sachs Research estimates that while data centres currently consume approximately 1-2% of the world’s overall power, this is expected to grow by 160% by the end of the decade, and 2022-2030 carbon emissions are predicted to double!
The rise in demand for data processing
The increase in artificial intelligence (AI) and machine learning (ML) must take responsibility for its part in this increased data consumption. The International Energy Agency reports that a Google search requires just 0.3 watt-hours of electricity, compared with a hefty 2.9 watt-hours for a single ChatGPT query. But while AI commands greater power requirements, it does have the potential to make data centres much more energy efficient.
By using AI algorithms to predict, monitor, and adjust power consumption in real-time, and optimise server utilisation and cooling systems, AI insights can minimise downtime by proactively addressing potential issues before they ever occur. From analysing disaster recovery scenarios to managing cooling systems by adjusting temperatures and airflow, AI modeling and simulation can leverage the complex trade-offs between performance, energy-efficiency and sustainability measures to establish optimal facility performance.
AI’s role in sourcing renewables
While prediction and adjustment of power consumption are important to the efficiency of the data centre, the benefits of AI adoption don’t stop there. Integrating renewable energy sources into operations, by using environmental energy sources, such as solar, wind, or water, provides a sustainable means of power that not only takes the pressure off the power grid but takes advantage of all the power given freely by our planet. AI can help by predicting the available resource’s power and production, aligning it with the data facility’s demands to reduce its carbon footprint. By evaluating alternative materials from renewable or sustainable sources for use in the structure, infrastructure, and building of a data centre, AI can help drive meaningful industry-wide change.
Shaping AI legislation and standards
There is a lack of standardisation within our sector. There are no global legislative policies and standards around sustainable AI development so there are no standards that responsible organisations can detail or present to their customers. Our industry must play a role in setting out what sustainable AI looks like, to determine the relationship between standards and policies on a national, international, and global basis.
We must have an informed understanding of the challenges our industry is facing, to allow us to assist in shaping policies and standards to drive the widespread adoption of sustainable AI. Conversation of real-world insights alongside technical expertise would enable practical and achievable guidelines to be implemented.
Associations and collaborations with industry partners, policymakers, industry bodies, and colocation facilities can help us by sharing the priority of sustainability and accountability up and down the supply chain.
The business advantage
While some organisations feel that offsetting a carbon footprint by planting a few trees allows them to claim sustainability, those who prioritise a truly sustainable future for development and innovation in the tech industry can reap real business advantages and tangible benefits.
● AI systems that improve operational efficiency and support the reduction of environmental impact can lead to lower operating costs offering significant cost savings. That can in turn translate to cost savings for their customers.
● By future proofing data centre infrastructure and containment systems to support emerging AI technologies, companies can better prepare for the growing power and cooling demands while maintaining a sustainable footprint.
● The top talent pool is increasingly prioritising sustainability. The next generation of AI experts seeks employers with strong environmental and social responsibility practices and credentials. Those organisations that do not place sustainability at the forefront of their business strategies could potentially lose out on recruiting the best.
The future of AI and sustainable data centres
While the use of AI demands ever-increasing high-performance compute processing power, data centres are likely to need to continue to increase their power consumption. Or do they?
AI is another step in the evolution of digital transformation. We can either allow it to consume more power or we can leverage it to enhance operational efficiency, integrate renewable energy sources and drive substantial and purposeful change. Many organisations will continue to embrace AI-driven solutions but, by developing and optimising the technology, it can pave the way for greener and more energy-efficient infrastructure, benefiting both business and planet.
A true 100% sustainable data centre doesn’t yet exist. Newly built data centres can make themselves as sustainable and efficient as possible, but it is the legacy data centres that will continue to have their work cut out. That's where new containment systems, ready for AI technologies such as on-chip cooling or the separation of hyperscale from standard processing in the same facility can enable companies to upgrade to more sustainable solutions, without having to completely rebuild or relocate.
Leveraging AI technology and taking full advantage of its benefits and influence on sustainability goals can pay huge dividends. By balancing the energy efficiency and cost savings rewards, combined with the not insubstantial marketing and promotional perks that provide added value, sustainable AI can deliver real business returns.
