With the rapid expansion of generative AI applications, data centre power demand is reaching unprecedented levels. According to a Cushman Wakefield report, AI-related data centre revenue is expected to increase 8,8 times by 2028 compared to 2024. However, AI servers consume five to ten times more power than traditional servers, especially during AI model training where energy consumption surges. This exponential demand raises critical concerns about whether traditional centralised power grids can continue to support AI-driven data centres. Nations and industries worldwide are facing increasing power scarcity challenges while data centres must also comply with energy efficiency and clean energy policies, further straining electricity supply.
Tech giants investing in diverse energy sources to secure power supply
To ensure a stable power supply for AI operations, major cloud and technology companies are making substantial investments in energy infrastructure. Microsoft and Google are actively exploring nuclear energy technologies to secure stable and efficient power sources, reducing their reliance on traditional power grids. Meanwhile, Amazon is investing heavily in clean energy initiatives, including large-scale solar and wind power projects complemented by energy storage solutions. These efforts aim to stabilise AI data centre operations during peak demand periods while mitigating risks associated with grid instability.
Transitioning from backup power to hybrid energy management
Traditionally, data centres relied on N+1 or 2N+1 backup power systems to ensure operational stability in the event of single or multiple failures. However, as AI workloads push power consumption to unprecedented levels and as diverse energy technologies emerge, data centres are shifting from traditional backup power strategies to hybrid energy management. By integrating advanced smart grids and microgrid technologies, data centres can dynamically allocate power resources, optimising energy efficiency. AI-driven energy management systems further refine power distribution strategies, reducing waste, enhancing operational performance and enabling more flexible energy deployment.
The future of AI data centres: Embracing hybrid energy and sustainability
Next-generation AI data centres will adopt more flexible and efficient power supply models. Hybrid energy infrastructures will become the standard, combining traditional power grids, renewable energy and energy storage solutions to ensure reliability while pursuing carbon neutrality. However, this transformation presents new challenges such as integrating multiple energy sources, enhancing power infrastructure monitoring, improving data communication efficiency and refining energy management strategies. As advancements in smart grids and clean energy technologies continue, AI data centres are poised to enter an era of high efficiency and low carbon emissions, providing a stable and sustainable power foundation for global AI computing.
Even in autonomous systems, human insight is important Schneider Electric South Africa
IT in Manufacturing
As AI, IIoT and digital twins reshape industrial operations, Schneider Electric’s Kobus Vermeulen and Bafana Msibi argue that the operator’s role is shifting from controlling equipment to interpreting intelligence, validating AI recommendations and making decisions that improve
Read more...AI-based quality inspection scales across P&G operations Siemens South Africa
IT in Manufacturing
Siemens and Procter & Gamble are expanding an AI-based visual inspection system across P&G’s manufacturing operations worldwide, inspecting products in real time at full line speed, and reducing scrap rates by 10 to 20% depending on the product.
Read more...How sustainability unlocks real business impact Rockwell Automation
IT in Manufacturing
Sustainability has moved from a compliance obligation to a strategic driver of resilience and value for manufacturers. Rockwell Automation explains how its Triple A framework and connected-data approach help producers cut energy, water and waste while improving productivity.
Read more...Mining’s AI future will depend on people as well as technology
IT in Manufacturing
South Africa’s mining industry is embracing digitalisation and AI, but AvenirHoldings chief executive officer, Tshidi Mndzebele says the real test is whether the people behind the industry are equipped to move forward with it, and outlines five things mining needs to get right.
Read more...How BIC overcame operational silos to integrate technology and people Rockwell Automation
IT in Manufacturing
Manufacturing giant BIC deployed Plex MES from Rockwell Automation at its Bizerte plant in Tunisia, replacing fragmented, partly in-house software with a single cloud-based execution system that broke operational silos and improved shop-floor visibility.
Read more...AI in industry is an early warning system, not a guarantee Schneider Electric South Africa
IT in Manufacturing
Schneider Electric’s Paul Steyn argues that AI in industrial maintenance is only as good as the infrastructure and people behind it, and explains the three layers a facility needs before predictive maintenance can deliver on its promise.
Read more...SA’s mining industry must put people at the heart of its digital future
IT in Manufacturing
South Africa’s mining industry employs around 470 000 people while its digital transition accelerates. Avenir’s chief executive officer, Tshidi Mndzebele argues that technology should enable and upskill the workforce rather than replace it, setting out five ways mining can make technology work for its people.
Read more...Quality control supports efficient tissue production Valmet South Africa
IT in Manufacturing
Valmet’s advanced quality control technology is helping tissue producers improve product consistency while reducing energy and raw material consumption. Its latest installation at KORO Paper in Vietnam combines real-time measurement, process optimisation and non-contact, nuclear-free quality control.
Read more...Patented algorithm maximises operator efficiency Rockwell Automation
IT in Manufacturing
Rockwell Automation piloted a patented anomaly detection algorithm and FactoryTalk DataMosaix at its Twinsburg electronics plant, giving operators 30 to 60 days’ advance warning of stencil failure, improving fail rates by 22% and achieving a 200% return on investment.
Read more...Cybersecurity must become part of the next era of industrial automation Adroit Technologies
IT in Manufacturing
The EU’s Cyber Resilience Act is setting a global benchmark for connected products, making cybersecurity a core engineering requirement in industrial automation. Adroit Technologies and Mitsubishi Electric explain how building security into controllers, PLCs and drives from the outset helps South African manufacturers protect production and meet international expectations.
While every effort has been made to ensure the accuracy of the information contained herein, the publisher and its agents cannot be held responsible for any errors contained, or any loss incurred as a result. Articles published do not necessarily reflect the views of the publishers. The editor reserves the right to alter or cut copy. Articles submitted are deemed to have been cleared for publication. Advertisements and company contact details are published as provided by the advertiser. Technews Publishing (Pty) Ltd cannot be held responsible for the accuracy or veracity of supplied material.
printer friendly version