Page 7 - Spring 2025 Electron

The rising Importance of High-density UPS Bridging the Gap between AI and Quantum
Systems in AI-driven Data Centres Computing in Data Centre Applications

This article, by Arturo Di-Filippi of Data Centre World This subject was addressed by Kalyan Kumar, Chief
exhibitor Vertiv, explains how AI is pushing data Product Officer for HCL Software, whose view is
centres toward higher density configurations in that combining quantum computing with AI will
order to accommodate increasing computational radically transform how data centres operate.
demands, linking this to the critical role played
DATA CENTRE TRENDS by uninterruptible power supply (UPS) systems in It is argued that the exponential growth in
addressing the challenges of densification:
data generation coupled with ever-increasing
computational demands is placing strain on
AND DEVELOPMENTS traditional architectures, and that quantum
computing and AI together are ‘transformative
forces’ that are capable of reshaping data centre
capabilities. Quantum computing can solve complex
problems beyond current capabilities, whilst AI
provides power to learn from vast datasets.

In unison, these technologies are envisaged to
redefine the operational paradigm of data centres,
The increasing Role of AI in Data Centres • Legacy Systems and Hybrid Approaches:
driving innovations in scalability, energy efficiency
Many facilities operate using legacy systems
Ramzi Charif, Vice President of Technical and computational performance:
that lack the capability to support advanced
Operations, EMEA for VIRTUS Data Centres, writing
AI solutions, but there is increasing adoption of ‘ In the context of data centres, quantum
in Data Centre Review, H1 2025, draws attention to
hybrid approaches that combine traditional computing offers groundbreaking advantages.
the increasing role of AI in data centre operations,
systems with AI enhancements. As an example For example, resource optimisation and workload
highlighting four specific instances:
existing sensors and control systems can ‘ AI-driven applications rely on high-performance distribution, both critical challenges in large-scale
The partnership aims to support up to 47,000 young be supplemented with AI algorithms to graphics processing units (GPUs) and central operations, can be approached through quantum
people to enjoy their own Big Bang at School day with provide advanced monitoring and predictive processing units (CPUs) that consume substantial algorithms which will improve the server efficiency
special science and engineering activities, workshops capabilities. amounts of power. To maximise efficiency and and reduce the overall operational costs.
and projects designed to arouse passion and curiosity capacity, data centres are densifying their
In his conclusion the author states: Certain experiments with quantum assisted
for all things related to STEM. The objective is to bring infrastructure, packing more computing power logistics optimisation demonstrated dramatic
tyhe same fun, excitement and interactive learning ‘ As AI technology continues to mature, its role into smaller footprints. reductions in computational times for routing and
experiences that are provided by the Big Bang fair to within data centres will likely expand. The next In densely packed environments, even minor scheduling tasks - an analogous challenge to
the classroom to benefit 11 to 14-year-olds. evolution could see AI not only managing disruptions in power delivery can cascade into managing server workloads in data centres. Such
operations but also integrating with external significant downtime or data loss. UPS systems advancements underscore quantum computing’s
• Cooling and Energy Optimisation: Cooling systems to provide real-time adjustments based designed for high density applications are vital for potential to redefine resource allocation, cooling
accounts for up to 40 per cent of all energy on broader network and environmental data. addressing these challenges. strategies and hardware deployment within data
costs in a data centre. Traditional methods This could include optimising power use across centres.’
have relied on fixed temperature settings and multiple facilities, integrating seamlessly with Energy efficient UPS systems are particularly
systems that cannot easily adapt to fluctuating smart grid technologies, and even coordinating essential for high density configurations, where Further benefits are envisaged in security with
loads or environmental changes. AI-driven energy use with renewable sources based on power demands are concentrated. By delivering quantum-safe cryptographic protocols, such as
cooling, however, is ‘changing the landscape’ availability and demand forecasts. power with minimal loss and enabling optimal lattice-based and multivariate encryption methods
load balancing, these systems support the high-
as machine learning algorithms monitor and
The industry’s acceptance of AI is expected to potentially providing increased resilience against
respond to temperature variations in real- density demands of next generation computing
grow, particularly as the technology demonstrates environments.’ quantum attacks in data centres.
time. AI systems optimise cooling dynamically,
its value through pilot projects and incremental
ensuring that resources are directed to where
improvements. The trajectory mirrors the It is noted that the unpredictability and variability
they are needed most.
industry’s shift towards cloud adoption: cautious
associated with AI workloads can put strain on
• Monitoring and Predictive Maintenance: at first, but eventually widespread as the benefits traditional power systems so advanced load
Data centres require constant surveillance become undeniable’. management capabilities are needed that can
of key metrics such as temperature, power readily adjust to sudden load changes.
consumption and network performance. AI Reference: Charif, R., ‘Could AI be crucial in helping Data Centres
optimise their own Operations?’, Data Centre Review, H1 2025, p.10-11
systems can provide this without the need for The author concludes:
human intervention. The ability of AI to analyse
‘ Densification, efficiency and modular solutions
vast amounts of data quickly means that it
can detect even minor deviations from normal are reshaping the future of data centre power
management. UPS systems are at the forefront of
performance.
this evolution, providing the reliability, scalability
• Workflow Automation: Routine tasks such as and energy efficiency needed to support the
performing equipment checks, monitoring demands of AI-driven workloads.’
systems and reporting can now be automated
using AI, improving consistency and reducing Reference: Electrical Engineering, February 2025, p.54-55
the risk of human error.

6 7
Become a member: https://membermojo.co.uk/ie Become a member: https://membermojo.co.uk/ie
Become a subscriber: https://membermojo.co.uk/subscriber Become a subscriber: https://membermojo.co.uk/subscriber
Become a sponsor: https://membermojo.co.uk/sponsor Become a sponsor: https://membermojo.co.uk/sponsor