Back to insights

Navigating the Generative AI Frontier

In the rapidly evolving data centre sector, the advent of Generative AI stands as a transformative force with far-reaching implications.  

Generative Artificial Intelligence (Generative AI) has emerged as one of the fastest-growing technologies in recent history. Still in its infancy, Generative AI has already demonstrated its potential through platforms like ChatGPT, which achieved unprecedented adoption rates. The platform acquired over 180 million subscribers within its first year.  

As businesses increasingly incorporate AI into their operations – leveraging predictive maintenance, resource allocation optimisation, and enhanced data analysis – the demand for robust data centres capable of supporting Generative AI workloads is rising faster than predicted. This will impose novel requirements on IT infrastructure, for which data centre operators need to be prepared.

The interplay between Generative AI & Data Centre Design 

Generative AI presents unparalleled demands that surpass the capabilities of conventional Cloud infrastructure. This necessitates a strategic re-evaluation of existing infrastructure to accommodate the evolving landscape of AI technologies and optimise their integration within data centre frameworks. Interestingly, the Equinix 2023 Global Tech Trends Survey (GTTS) found that 42% of IT leaders express a lack of confidence in their infrastructure’s capability to effectively support the expanding utilisation of AI. 

Meta is one organisation that has taken a proactive approach. In response to the transformative impact of AI on the data centre landscape, the company has undertaken a significant redesign of its data centre projects worldwide. This strategic shift involves scrapping existing projects and embracing a novel approach defined by the first of its next generation data centres in Temple, Texas. 

Meta has acknowledged CPU limitations in AI tasks and has shifted to GPUs and accelerators, implementing hybrid liquid and air cooling for flexibility, streamlined power distribution and enhanced scalability. The redesign targets a 31% cost reduction and faster construction, promising an efficient data centre infrastructure for the AI era. 

Transforming Data Centre Models with Generative AI 

Data centres can leverage Generative AI to revolutionise their own operational models. For instance, Generative AI can be applied to: 

  • Predict workload patterns  
  • Optimise resource allocation 
  • Reduce energy consumption 
  • Lower operating expenditure 
  • Improve sustainability 
  • Analyse data centre infrastructure to design more reliable facilities 
  • Optimise layout options for new data centres 

Additionally, a comprehensive approach to the lifecycle management of data centre assets, including Mechanical, Electrical, and Plumbing (MEP) equipment and servers, can be facilitated through Generative AI tools for asset optimisation, extending their operational lifespan, and positively impacting costs and sustainability. This may include reducing the capital expenditure required for frequent equipment replacements, leading to substantial cost savings and a more efficient allocation of resources. 

Generative AI could also provide insightful analysis on the operating and capital costs of a data centre, enabling informed decision-making in all areas of development, from site selection and master planning to design development, construction management and operations. 

Generative AI’s role extends to environmental considerations, such as assessing cooling opportunities and the status of nearby grid infrastructure, as well as exploring options for reusing waste heat in surrounding entities like farms, fish farms and schools.  

Adopting a holistic approach has the potential to minimise the overall carbon footprint of data centre operations, showcasing the potential for new software tools developed from a Generative AI perspective to drive innovation and efficiency in the evolving landscape of data centre management. 

The Ethics of Artificial Intelligence 

AI is a relatively new entrant with substantial potential to revolutionise the industry – and the world. It is imperative to exercise conscientiousness regarding the data generated and the tools employed. Questions surrounding the extent of control delegated to these tools and their societal impact demand careful consideration. 

While Deerns advocates for technological progress, we also acknowledge the challenges and nuances in its application and influence. Guidelines from governments and businesses are eagerly awaited to navigate this evolving terrain. 

In the recently unveiled Bletchley Declaration, countries recognise the vast opportunities presented by AI to enhance human well-being, peace, and prosperity globally. Emphasising the need for AI to be safe, human-centric, trustworthy, responsible, aligned and transparent, the declaration affirms the importance of international cooperation to address challenges and risks posed by AI. 

Navigating the Generative AI Frontier

Here’s where we navigate the Generative AI Frontier  

With a commitment to awareness, collaboration, and a forward-looking stance, Deerns is poised to be a driving force in the evolution of data centre design and operations. As technology evolves and challenges emerge, Deerns remains at the forefront, engaged with the supply chain, conducting in-depth research, and participating in networking initiatives.  

In these exciting times, Deerns’ dedication to innovation and sustainability signals a promising future, where the synergy between technology and responsible practices paves the way for a new era of data centre excellence.

Let’s talk

Jonathan Clark

Data Centre Sustainability Lead