Pan African Data Centres and Cloud Solutions Africa host webinar. This article is part five of a six-part series.
There is no one-size-fits-all solution when it comes to integrating artificial intelligence into existing infrastructure. DC Studio | Freepik.com
In a bid to explore the burgeoning role of Artificial Intelligence (AI) in shaping the future of data centres, the Pan African Data Centres and Cloud Solutions Africa jointly organised a webinar titled ‘AI – The driving force of future data centres’. The virtual event delved into the remarkable advancements in data centre density and the pressing need for energy efficiency and sophisticated cooling solutions.
The webinar, hosted by Amy Hopkinson, conference producer of the Pan African Data Centres 2024 event, featured a distinguished panel of industry experts:
- Vanessa Moffat, EMEA Channel Partner manager at EkkoSense
- Samuel Chukwukereuba, Data Centre manager, Project Implementations at Digital Realty
- Niraj Shah, Sales and Business Development director at IX Africa
- Keith Sullivan, director of Strategic Innovation at AFL
As the conversation delved into the implications of AI deployment on existing data centre infrastructure, the panellists provided valuable insights into the challenges and opportunities for retrofitting colocation facilities to accommodate the evolving demands of AI workloads.
Keith Sullivan highlighted the scalability limitations of current colocation data centres in sub-Saharan Africa, noting their typical capacity of a few megawatts. He emphasised the need for meticulous design considerations, particularly in terms of load-bearing capacity and efficiency metrics such as power usage effectiveness (PUE). Sullivan underscored the importance of adopting efficient cooling solutions to mitigate heat dissipation challenges, acknowledging the complexities involved in retrofitting existing facilities to meet high-density AI requirements.
In response to a listener question regarding the feasibility of retrofitting existing data centres for AI deployment, Sullivan emphasised the difficulties faced by colocation facilities in accommodating high-density racks and heavier equipment. He noted the importance of adhering to industry standards, such as the Open Compute project guidelines, to facilitate seamless integration of AI technologies into existing infrastructure.
Despite the challenges posed by retrofitting, Sullivan highlighted the potential for Edge data centres to serve as AI inference locations, catering to smaller-scale deployments in remote or distributed environments. He emphasised the diverse applications of AI across various industries, underscoring the need for flexible and adaptable data centre solutions to meet evolving demands.
In conclusion, the panellists recognised the complexities involved in retrofitting existing data centres for AI deployment, particularly in terms of scalability, efficiency and cooling requirements. While challenges exist, they emphasised the importance of adopting industry best practices and leveraging emerging technologies to optimise infrastructure and unlock new opportunities for innovation in the AI era.
In a recent discussion among industry experts, the conversation revolved around the nuanced nature of AI deployment in African data centres. As technology continues to evolve, it’s become clear that there is no one-size-fits-all solution when it comes to integrating artificial intelligence into existing infrastructure.
