Innovation and Impact: The AI Sustainability Paradox
As AI adoption surges, data centres are having to evolve in order to prioritise sustainability.
The growing energy demands of AI is starting to present a significant challenge. Training a single AI model can emit tonnes of carbon dioxide, which is very significant when considering the number of data centres operating worldwide.
To address this, the data centre industry has started developing strategies to be more energy efficient. This requires significant collaboration with stakeholders and higher levels of computing innovation.
The path forward lies in balancing AI's potential with its environmental costs.
Some technology leaders, however, are advocating for AI’s ability to actually improve sustainable progress. Kyndryl, a major player in the IT infrastructure services market, is just one company that is leveraging AI to optimise data centre operations and improve sustainability.
Making the most of AI
Nick Drouet, Chief Technology Officer for Kyndryl UK&I, is responsible for how the company tackles the most complex challenges Kyndryl customers face.
“AI and machine learning (AI/ML) are fundamentally reshaping the data centre market, driving efficiency, automation and sustainability,” he shares. “These technologies are not only optimising operations but also redefining how data centres are designed, managed and scaled.”
Kyndryl’s Data Center Advisor platform harnesses AI/ML to analyse operational data, whilst also providing predictive insights to enhance reliability and efficiency to reduce costs.
“Kyndryl is actively enhancing its position in the AI/ML market by integrating these technologies into its service offerings and operations,” Nick says. “To scale AI/ML technologies in support of the data centre sector, Kyndryl is focusing on modernising its facilities and forming strategic partnerships.
“To achieve this, we’re transitioning to more efficient data centres that consume less energy, leveraging AI to optimise cooling systems and manage energy consumption effectively.”
The company also recently expanded its partnership with Nokia to offer advanced data centre networking solutions.
Nick shares: “This collaboration combines Kyndryl's expertise in designing and managing hybrid cloud environments with Nokia's high-performance networking infrastructure, aiming to deliver scalable and secure data centre solutions for enterprise customers.”
The journey towards AI data centres
As businesses transform and become more data-focused, their IT infrastructure will need to evolve. This, Nick explains, inevitably changes the nature of their data centre requirements, which leads to all kinds of challenges.
“While most corporations will have a digital transformation strategy in place, few will have a data centre strategy. This needs to change if they’re to remain competitive,” he says.
“As AI adoption accelerates, data centres will continue evolving — becoming more intelligent, autonomous and sustainable. Organisations that leverage AI early and effectively will not only drive operational excellence but also future-proof their infrastructure in an increasingly digital world.”
Some of the benefits AI can offer a data centre include:
- Operational automation
- Predictive analytics and predictive maintenance
- Reducing downtime
- Incident prevention
- Lowering cost
- Optimising resource allocation
The critical challenge remains: How can data centre operators harness AI/ML to help, rather than hinder, the sustainability of their data centres?
Nick suggests that the technologies can help by optimising energy use, reducing waste and improving efficiency.
“One of the most impactful applications is AI-driven cooling management, where ML algorithms analyse real-time sensor data to adjust cooling systems dynamically, significantly reducing energy consumption,” he says. “AI-driven optimisations contribute directly to lower power usage and higher operational efficiency.”
AI can also help data centre operators optimise their workloads and integrate renewable energy like solar and wind power more effectively, whilst distributing workloads across servers to prevent energy waste.
“AI can enhance energy storage management, ensuring that green energy is efficiently utilised, further reducing reliance on carbon-intensive power sources,” Nick adds. “It can also provide real-time insights into a data centre’s energy usage, water consumption and emissions to allow operators to improve sustainability reporting, meet net-zero targets and align with industry regulations.
“By leveraging AI/ML, data centre operators can drive cost savings, operational efficiency and environmental responsibility, making AI a critical tool for the future of sustainable data centres.”
To read the full article in the magazine, click HERE.
Explore the latest edition of Data Centre Magazine and be part of the conversation at our global conference series, Tech & AI LIVE and Data Centre LIVE.
Discover all our upcoming events and secure your tickets today
Data Centre Magazine is a BizClik brand