The Data Centre Dilemma: Keeping Up With AI Innovation
Enterprises around the world investing in AI applications have left data centre operators scrambling to confront its high levels of processing power and energy consumption.
As the technology continues to grow rapidly, contributing to substantial data centre investments, Digital Realty cites that data centre energy consumption is expected to increase by 12% by 2030. This is due to resource-hungry workloads that require more power and cooling systems to run safely and effectively.
The influence of AI applications can no longer be ignored within the data centre industry. As the global market is expected to swell to meet growing demands for workloads and storage needs, those within the industry are urgently calling for change.
"We need to be prepared for these changing demands on our infrastructure, while prioritising scalability and efficiency, at all times,” comments Brandon Tikalsky, Vice President, Technology Strategy & Business Operations at Digital Realty. “It’s our job to make sure businesses can continue to embrace change in a sustainable way.”
AI is here to stay: Adapting to a new landscape
As AI and Generative AI (Gen AI) workloads continue to become more widespread, the management and intensity of these workloads pose the biggest challenge for the data centre industry: particularly when it comes to environmental sustainability.
With this in mind, it has been suggested the design of data centre facilities requires rethinking, as current infrastructure is not sufficient to cope with future demand.
“In 2024 and beyond, AI is poised to play a pivotal role in reshaping how we design, construct, and manage data centres,” highlights Sam Bainborough, Sales Director EMEA at Vertiv. “Virtually every industry is exploring opportunities to drive business value through AI, and this is posing several questions and challenges on how to best deploy the required infrastructure in a timely and efficient manner.”
As hyperscale data centres become more prevalent, there is a greater need for optimising storage space and performance through high-density server racks. According to a report from DataCenter Knowledge, in 2016 average rack density was sitting around 6.1 KW. In 2024, this has now more than doubled to 12 KW.
And with a report from Synergy finding that hyperscale data centre capacity will nearly treble in the next six years directly as a result of AI growth, ensuring data centre energy efficiency is vital.
“A notable trend is an increasing focus on energy sustainability, with data centres leading the charge in adopting renewable energy sources and implementing energy-efficient solutions like liquid immersion cooling to reduce environmental impact,” comments Barath Narayanan, Global BFSI and Europe Geo Head at Persistent Systems. “This transformation demands more than just incremental improvements; it calls for a complete rethinking of data centre architecture.”
Sustainability in focus: Keeping data centres cool
As AI continues to evolve, it will present both opportunities and challenges within the data centre industry - and ultimately shape its future. It will be vital for data centre businesses to adapt to these changes in order to remain competitive and be successful in the long term.
Notably, AI and high-performance computing (HPC) will continue to disrupt the data centre landscape because current data centres are not designed to support the high power requirements of AI racks.
“As a result, alongside this AI transformation, data centre providers are facing an unprecedented demand for computing power and seeking alternative energy sources,” states Sam. “Adapting to this situation demands that data centres adopt creative solutions that are capable of delivering increasing computing power, whilst employing energy-efficient methods. This includes expanding the use of alternative energy, smart grids, hybrid grids and innovative data centre designs to deliver reliable solutions for customers, while lessening the negative impacts on our planet.
“By prioritising power efficiency, data centres can not only tackle the demands posed by growing AI workloads but also contribute to a more environmentally conscious and sustainable future.”
The rise of AI is also changing HPC in data centres, meaning that the data centre industry must look at how it can keep facilities cool for optimum performance.
“The rise of specialised AI processors providing the required computing power has only heightened the need for energy efficiency, and alternative sources such as solar and wind power should be used to meet the growing demand for more sustainable electricity generation and consumption,” states Sam. “These options not only provide environmentally friendly solutions but also contribute to grid stability and enable efficient resource use.”
Addressing growing demands for data centres
As the data centre industry confronts the AI challenge, it must tackle myriad issues to continue achieving growth.
Notably, capacity is one of the largest issues facing the data centre industry today. With facilities running at higher power densities, data centre operators are crying out for more advanced cooling solutions to support their rack configurations.
“Adapting data centre design and infrastructure to accommodate these new technologies and cooling methods will have significant capital expenditure implications going forward,” highlights Rajesh Sennik, Partner at KPMG, who explains how AI is significantly driving demand for data centre capacity.
“Initially, the jump in demand relates to training these models, but over time, this will shift to inference, where the model can draw conclusions from brand-new data,” he adds. “This will require more Edge data centre capacity as it allows data processing to be closer to where it is needed, which can reduce latency and improve response times in real-time applications.”
With the influence of AI – and all that comes with it – data centres are evolving to become real drivers of innovation. However, this means that these facilities must adapt to satisfy higher demands.
“As power usage grows, sustainability and environmental impact is an important concern with multiple companies putting a premium on GreenOps,” Barath notes.
GreenOps refers to the balance between operational and financial efficiency, with businesses striving to achieve the lowest environmental impact possible.
“Data centres are exploring renewable energy sources like solar, wind, geothermal and wave energy, employing AI-driven analytics to optimise operations for energy efficiency and reduced carbon footprint,” Barath adds. “The industry is evolving to meet these challenges, ensuring data centres can support the burgeoning AI ecosystem.”
Another solution is to construct AI-ready data centres, which could confront rising storage demands and the pressing needs of AI technologies, as Brandon explains.
“Innovation breeds further innovation, and there’s no doubt that Gen AI will continue to shape the data centre industry. We need to make sure sustainability is a consideration for all future innovations we introduce as an industry – from renewable energy adoption through to responsible resource management.
“We need to keep pursuing and exploring new AI-driven initiatives, innovations and partnerships to enhance operational efficiency even further.”
******
Make sure you check out the latest edition of Data Centre Magazine and also sign up to our global conference series - Tech & AI LIVE 2024
******
Data Centre Magazine is a BizClik brand