AI can only reach its potential if data centres evolve to accommodate it

By Sam Bainborough, Director EMEA-Strategic Segment Colocation & Hyperscale at Vertiv.

From personalised shopping experiences to smart transport systems, Artificial intelligence (AI) is becoming more and more ubiquitous in our lives, fundamentally changing how we live and work. And with the arrival of ChatGPT and other AI driven apps, it is clear that this technology is reshaping the digital landscape and redefining what data centres need to handle. Many industry experts are anticipating the ongoing growth of AI, including Statista who predicts that the market will reach US$305.9bn in 2024 and expand to US4738.80bn by 2030. 

However, AI will only be able to succeed if the technology infrastructure evolves to accommodate the computing power required to process the data and run the IT systems that enable this growing phenomenon. This means that data centres are crucial in enabling this technological revolution. 

The evolution of data centres

To meet the escalating demand for computing power, data centre operators are prioritising the development of infrastructure capable of not only satisfying current needs but also accommodating future growth. They must rethink design strategies and make significant changes to network architecture, power systems and thermal management. A holistic approach is required emphasising adaptability, energy efficiency and reliability - not least because data centre owners are confronted with multiple challenges such as carbon reduction regulations, surging power requirements and heightened heat generation. 

The combination of increasing bandwidth requirements and the continuous influx of data is fundamentally reshaping the business landscape, and prioritising efficiency and innovation is even more paramount. Two key areas that need to be addressed are power, to cope with the demands of AI workloads, and thermal management, to enable the critical digital infrastructure to operate as efficiently as possible.

Power 

High Performance Computing (HPC) is changing with the rise of AI, causing a significant increase in power demands, fuelled by the adoption of specialised processors essential for managing complex tasks. Typical data centre computer racks are expected to increase from 5 kW to 7 kW today (equivalent to the size of a small residential backup generator) to 50 kW or more in the not-too-distant future, according to Omdia’s 2022 Data Center Thermal Management Market Analysis report. Addressing this challenge demands that data centres adopt creative solutions for managing and optimising power usage. Integrating energy-efficient hardware into data centres requires implementing state-of-the-art hardware designs and staying updated with advancements in processor technology. 

The strategic emphasis on power efficiency aligns with the broader imperative of promoting sustainability in the face of escalating energy consumption. 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. 

Thermal management and cooling solutions

To maintain optimal performance and prevent hardware failures due to overheating, data centre operators keep abreast of innovative cooling technologies. Liquid cooling systems are becoming more popular, perfectly integrating with chilled water systems that are an established choice. Striking a delicate balance between processing power and thermal management is crucial, safeguarding the longevity and reliability of the entire data centre infrastructure.

There are three liquid cooling technologies that are in use today:

Air-to-liquid cooling: While these technologies don’t bring liquid directly to the server, they do utilise the high thermal transfer properties of liquid. Passive or active heat exchangers replace the rear door of the IT equipment rack with a liquid heat exchanger. With a passive design, server fans expel heated air through a liquid-filled coil mounted in place of the rear door of the rack. It is the coil that absorbs the heat before the air passes into the data centre. Active heat exchangers include fans to pull air into the datacenter through the coils and remove heat from even higher density racks. These systems can be used in conjunction with air-cooling systems to cool environments with mixed rack densities. 

Direct-to-Chip cooling: This is where a cool liquid is circulated to cold-plate heat exchangers embedded in the IT equipment. With this method, servers and chips are integrated with micro fluid channels close to the heat generating components. The heated fluid is transferred outside the rack for heat rejection, which is typically performed by the CDU. Direct-to-chip cooling technologies generally have higher heat removal capacities than rear door heat exchangers.

The next phase in this evolution is the immersion cooling.

Immersion cooling: This involves submerging servers and other components in a thermally conductive dielectric liquid or fluid. With this method, the need for air cooling is eliminated, including the fans within servers. This approach maximises the thermal transfer properties of liquid and is the most energy efficient form of liquid cooling.

Can liquid cooling and air cooling co-exist? 

Air-cooled and liquid-cooled solutions will absolutely co-exist and data centres will need to tightly orchestrate both to optimise the overall environment within the facility. Even within liquid-cooled servers, air cooling will continue to be required as direct-to-chip technology  does not remove 100% of the heat from equipment in the rack. Typically, this approach can remove 70-75% of the heat generated by the equipment in the rack, requiring a hybrid approach to cooling. 

It's important to note that innovative thermal management practices not only optimise performance but also reduce the environmental impact of data centres, emphasising energy efficiency and resource utilisation. By minimising energy consumption and maximising resource utilisation (including waste heat recovery to support the circular economy), sustainable cooling practices play a pivotal role in mitigating the ecological impact of data centre operations.

Ensuring a Holistic Approach

Successfully navigating this evolving dynamic landscape necessitates a holistic approach. One of the keys to success lies in involving all stakeholders, recognising the importance of collaboration and communication across diverse disciplines. Engaging not only power and cooling specialists but also those responsible for facility management, storage and technology deployment, fosters a comprehensive understanding of the data centre’s intricate requirements.

As data centres evolve with denser configurations and rapid technological advancements, design must go beyond technical specs. It involves efficient decision-making and proactive engagement with industry experts to navigate the complexities of AI-driven transformations. Holistic design helps with streamlining decision-making processes while considering lead times and involving stakeholders at every stage.

---/---

AI is a fascinating technology that’s poised to change the world, but it’s impossible to predict exactly how it will evolve and what it will do. However, its potential is only as great as the world’s data centres’ capacity to support the computational intelligence it will require. The data centre industry must continue to evolve to provide the dynamic and innovative cooling and power solutions needed to support data centre challenges and maximise AI’s true potential.

This proactive approach not only enables data centres to meet the burgeoning demands of AI applications but also positions them as catalysts for progress. By prioritising innovation, data centres can mitigate their environmental impact as well as enhance their resilience in the face of evolving challenges such as climate change and resource scarcity. Moreover, by optimising efficiency through innovative practices and technologies, data centres can operate at optimal performance while minimising energy consumption and operational costs.

Ultimately, the significant AI proliferation heralds a new frontier in data centre design. By embracing a holistic approach, data centres are poised to lead the charge in innovation, spearheading progress in an era defined by growth and technological advancement.

By Ian Ray, Head of Date, AI and ML, Daemon.
Despite the massive opportunity presented by AI for the services industry, 54% of services leaders...
By Dael Williamson, EMEA CTO at Databricks.
By Shadi Rostami, SVP of Engineering at Amplitude.
By Gregg Ostrowski, CTO Advisor, Cisco Observability.
By Rosemary Thomas, Senior Technical Researcher, AI Labs, Version 1.