Schneider Electric in a recent paper titled, ‘The AI Disruption: Challenges and Guidance for Data Centre Design’, delved into the crucial attributes and trends surrounding Artificial Intelligence (AI).
The paper sheds light on the resulting challenges faced by data centres and offers comprehensive guidance to address these issues in power, cooling, racks, and software management.
According to the white paper, the data centre industry was set to face significant tasks in 2024, as it adapts to the demands and harnesses the potential of AI.
Managing scarcity and power-related issues will be paramount, accompanied by the emergence of three key trends: alternative energy, liquid cooling, and quantum computing.
“The surge of AI in data centre workloads, from large training clusters to small edge inference servers, signifies a shift towards higher rack power densities. AI start-ups, enterprises, colocation providers, and internet giants are urged to consider the impact of these densities on the design and management of data centre physical infrastructure,” it noted.
Schneider Electric predicted that AI would play a pivotal role in enhancing asset management for operators. The impact will extend to various aspects, including utility and backup power control, cooling control, operational optimisation through digital twins, data centre design and construction, maintenance, robotics, and more.
The white paper emphasised the continued adoption of cloud computing and digital transformation by enterprises. However, it identifies AI as the leading driver of data centre demand in the coming year, projecting three AI data centers for every new standard data centre, each offering higher server densities.
Addressing the challenges of scarcity and modularity, the white paper suggested that modular data center infrastructure is a viable solution. Purpose-built facilities still require components, leading operators to secure the supply chain as much as possible.
The critical issue of power and grid sustainability was also discussed. Operators were urged to consider renewable energy sources and large battery storage systems to meet ambitious goals for carbon neutrality.
The company identified three additional trends: alternatives to standard diesel, liquid cooling options, and the emergence of quantum computing. Green diesel and green hydrogen are suggested as sustainable alternatives, while liquid cooling becomes crucial for AI clusters exceeding 20 kW/rack. Quantum computing, although not anticipated to move into typical data centres this year, prompts the need to protect infrastructure from potential cybersecurity compromises.
As the industry grapples with the aftermath of the AI surge, Schneider Electric emphasizes the importance of tapping into its extensive resources for gaining clarity in the face of challenges such as data center shortages, grid issues, and more. The white paper serves as a valuable guide for navigating the evolving landscape of data centre design in the era of AI disruption.
