When data converged with medical and life science research, it created a new discipline: bioinformatics – transforming how scientists understand and study the building blocks of life.

Today, researchers from across the world use technology to interpret vast biological datasets, helping accelerate scientific discovery for the betterment of humankind.
At the centre of this transformation is high-performance computing (HPC) a subset of advanced computing providing the infrastructure and processing capability needed to train and run Artificial Intelligence (AI) and deep learning models.
From bioinformatics to genomic sequencing and diagnostic imaging, HPC and AI are becoming increasingly interconnected with research. But as computing requirements grow, so too do the infrastructure challenges connected to them – in particularly power, cooling, rack densities, space and resilience.
At the same time, another transformation is underway. New advances in HPC and machine learning have accelerated scientific discovery at unprecedented pace. From identifying new drug candidates and modelling disease pathways, to enabling precision medicine and AI-assisted diagnostics, computational power, and the data centres that provide it, have become as critical to research as the laboratories studying it.
Behind every breakthrough sits a growing infrastructure challenge. As demand for computing power increases, Life Science and research organisations must balance performance, resilience, sustainability and scalability, while ensuring researchers remain focused on what matters most: advancing science and improving outcomes, no matter for whom.
The new reality: more compute, greater density and limited space
Today AI-powered workloads require dense clusters of graphics processing units (GPUs) and CPUs, supported by vast storage and low-latency networking. At the same time, genomic sequencing, medical imaging and digital pathology continues to generate exponentially larger datasets, requiring more space, and more processing power than ever before.
For laboratories and research institutions, the challenge extends far beyond simply adding more compute capacity. Many campuses require space for experimental and scientific work, while increasingly powerful computing workloads demand additional power, cooling and operational oversight. At the same time, IT and data centre teams must address some of their most pressing challenges, including scaling infrastructure to support growing workloads, maintaining resilience for mission-critical research, protecting sensitive scientific and data, and improving coordination between IT and facilities teams to ensure power, cooling and capacity planning remain aligned. These priorities become even more complex as researchers increasingly depend on real-time access to data and AI-driven insights.
The location of the compute, which must often be placed in close proximity to the sequencing equipment adds another layer of complexity due to the levels of low latency, or speed of data transfer required. This means that genomic sequencers, medical imaging systems and advanced testing platforms producing vast datasets must process digital information closer to where the data is generated and consumed, before transferring it to centralised data centre environments.
The question, therefore, is no longer simply adding more compute. It is how to do so efficiently, resiliently and at-scale, and without compromising the primary purpose of these facilities.
Data centre infrastructure must evolve with scientific computing
As HPC and AI workloads continue to grow, traditional approaches to infrastructure are reaching their limits. Life Science organisations, for example, require technologies that can scale alongside research demands without disrupting critical operations.

Prefabricated, modular data centre solutions provide a flexible path forward. Enabling organisations to expand computing capacity without consuming valuable clinical or research space. Modules can be deployed externally and expanded rapidly supporting new digital initiatives, while allowing them to consider both current and future capacity requirements, and changing external environmental conditions.
Cooling must also evolve to accommodate increasing rack densities, with both AI clusters and GPU-intensive workloads placing unprecedented demands on thermal management, making liquid cooling imperative for rack densities of 36 Kilowatts (kW) or more.
Computational Fluid Dynamics (CFD) is also vital – helping identify hotspots and airflow issues before physical changes are made – and at the same time, greater operational visibility is equally important.
Data Centre Infrastructure Management (DCIM) platforms, such as Schneider Electric’s EcoStruxure IT software, provide real-time information about infrastructure, helping operators make informed decisions around power, cooling and capacity while moving from reactive to predictive operations.
Where Data Centres meet Scientific Discovery: HPC and Research in practice
These challenges are already shaping the data centre infrastructure strategies of many of the UK’s leading research organisations. At the Wellcome Sanger Institute, for example, genomic research produces vast quantities of sequencing data, and its data centre consumes 4MW of power with more than 400 racks.
Critical to its research is the resiliency, visibility and operational efficiency of its infrastructure, and working with EcoXpert Partner EfficiencyIT, the Institute deployed Schneider Electric’s EcoStruxure™ IT Expert to gain greater visibility and support the management of its highly data-intensive research environment – resulting in a reduction of its energy consumption of over 35%.
The Pirbright Institute is another world-leading centre for research into diseases of farm animals and viruses that spread from animals to humans, and faced a different challenge. Working with Schneider Electric and its EcoXpert Partner, Advanced Power Technology (APT), Pirbright deployed a containerised data centre designed to support new developments in HPC, AI and GPU-powered computing.
The new facility provides 80kW of scalable capacity in an N+1 configuration. Importantly, the modular approach enabled Pirbright to deploy new digital infrastructure away from its existing buildings while supporting the resilience, security and continuity requirements of its scientific research.
Advancing sustainable research
While many research institutions are now focused on enabling scientific discovery, others are also examining how their computational infrastructure can contribute to wider sustainability objectives – an opportunity which extends far beyond traditional efficiencies.
In the higher education sector, organizations like Queen Mary University of London are exploring opportunities to recover and reuse waste heat generated by data centres, and use it to generate energy for campus district heating and local hot water systems.

Delivered in collaboration with Schneider Electric and its EcoXpert Partner, Advanced Power Technology (APT), the project uses a multi-stage heat recovery process to transform waste heat into water at temperatures of 65–75°C – demonstrating how organisations can consider the data centre as part of a wider energy ecosystem, rather than viewing them solely as consumers.
Next-generation data centres fuel tomorrow’s breakthroughs
While HPC is becoming fundamental to research, scientific progress cannot be separated from the infrastructure enabling it, meaning breakthrough research requires equally advanced data centre infrastructure to support it.
By planning for future capacities, adopting scalable, modular solutions, addressing cooling strategically, and using software to improve visibility, organisations can embed sustainability into infrastructure planning, and create resilient HPC environments capable of evolving alongside research.
At Schneider Electric, our experience supporting some of the UK’s most advanced research organisations demonstrates that successful HPC deployments require more than computing power alone. They depend on an integrated approach to power, cooling, software, sustainability and lifecycle management and a skilled ecosystem that can deliver it.
To learn more about our end-to-end AI-ready data centre infrastructure solutions, advanced cooling technologies and strategies for AI data centre cooling, visit Schneider Electric’s AI Factory hub.
Add a comment