{"id":170823,"date":"2026-08-19T05:49:19","date_gmt":"2026-08-19T04:49:19","guid":{"rendered":"https:\/\/www.intelligentcio.com\/eu\/2026\/08\/19\/accelerated-compute-and-the-heat-reuse-opportunity\/"},"modified":"2026-09-22T12:56:50","modified_gmt":"2026-09-22T11:56:50","slug":"accelerated-compute-and-the-heat-reuse-opportunity","status":"publish","type":"post","link":"https:\/\/www.intelligentcio.com\/eu\/2026\/08\/19\/accelerated-compute-and-the-heat-reuse-opportunity\/","title":{"rendered":"Accelerated compute and the heat reuse opportunity"},"content":{"rendered":"\n

Schneider Electric\u2019s Matthew Baynes, VP, Secure Power & Data Centres in the UK & Ireland, explores how turning waste energy from AI and high performance infrastructure can be a net zero advantage.<\/em><\/p>\n\n\n

\n
\"\"
Matthew Baynes, VP, Secure Power & Data Centres in the UK & Ireland<\/em><\/figcaption><\/figure><\/div>\n\n\n

There is an old bit of wisdom that asserts in every problem there is an opportunity.<\/strong><\/p>\n\n\n\n

Accelerated compute and AI adoption are driving a significant build out of increasingly sophisticated digital infrastructure systems \u2013 all of which feature high-density racks and servers, increased energy demands and the transition towards hybrid and liquid-cooled systems, at a time unlike anything seen before.<\/p>\n\n\n\n

Whitespace equipment, for example, has become heavier with implications for racks and row architectures and GPU configurations. Floor loading has become a key consideration for deploying GPU-accelerated servers and power demands have gone from 40-60 kW per rack to 100+, with densities expected to reach 1MW in the near future.<\/p>\n\n\n\n

At the same time, hyperscalers, neoclouds and colocation operators have embraced new cooling techniques to cope with increased thermal demands, and direct to chip liquid cooling, combined with air-cooled servers, is now the leading cooling topology.<\/p>\n\n\n\n

This change, combined with the energy and resource demands of AI, has also impacted sustainability efforts, with many operators left struggling to meet both regulations and environmental goals. Amid the challenge, there is now a growing opportunity to not only prioritise sustainability, but to become active energy assets via heat reuse.<\/p>\n\n\n\n

What is data centre reuse?<\/strong><\/p>\n\n\n\n

In years gone by and as data centre equipment densities grew, new cooling techniques were developed to cope with the increased volume of heat produced by servers and CPUs (Central Processing Units). Hot and cold aisle containment configurations, along with more efficient heat extraction, allowed owners and operators to keep their equipment within a specified window of optimal operation.<\/p>\n\n\n\n

However, despite recognising there was value in the heat extracted, heat reuse was technically challenging and with uncertain reward. In the UK and Ireland, for example, many operators now build data centres that include heat reuse capability but have nothing to connect to at the boundary \u2013 limiting the ability for data centres to become contributors to local communities. In Europe, the advent of greater regulatory supervision via the European Energy Efficiency Directive (EED)<\/a> has also refocused the industry on the issue but highlighted the potential opportunity.<\/p>\n\n\n\n

Why heat reuse matters<\/strong>  <\/p>\n\n\n\n

Data centres consume significant amounts of electricity and much of that energy ultimately becomes heat. In fact, the IEA<\/a> predicts data centres will represent around 3% of total global electricity consumption by 2030.<\/p>\n\n\n\n

Historically, operators focused on removing this heat as efficiently as possible. Today, the industry is increasingly recognising that waste heat can be repurposed as a valuable energy source and can be a valuable part of their efforts to achieve net zero.<\/p>\n\n\n\n

For 2024, the State of European Data Centres report<\/em><\/a> found that 31% of data centres had an actively measured Energy Reuse Factor (ERF) fully supported by data. For 2025, this had risen to 39%. The report also found that the greatest challenge for heat reuse, distance from heat distribution networks, had fallen by 15% from the previous year and uncertainty on legal responsibility and requirements had fallen by 13%.<\/p>\n\n\n\n

Putting heat reuse into practice<\/strong><\/p>\n\n\n\n

A great example of heat reuse in practice can be seen at Queen Mary University of London<\/a> where Schneider Electric, in collaboration with its partners at Advanced Power Technology (APT), undertook a data centre modernisation programme<\/a> to increase high-density computing functionality, upgrade the university\u2019s inefficient cooling and increase reliability and availability issues, while implementing heat recovery.<\/p>\n\n\n\n

Using a Hot Aisle Containment System (HACS) and\u00a0Data Centre Infrastructure Management software, the companies were able to deliver a data centre with enhanced High-Performance Compute (HPC) capabilities. Recovered heat began to be used for heating and hot water across the main campus and the data centre now meets both current and future research computing needs, while supporting sustainability.<\/p>\n\n\n\n

From a more technical perspective, the university\u2019s heat reuse system leverages the HACS and feeds heat pumps and heat exchangers, providing 75c water for both storage and supply for heating. The system, for example, can provide 2.8 million back-to-back showers for students, up to 52 simultaneously, delivering a valuable resource for the institution.<\/p>\n\n\n\n

Heat reuse at larger scale<\/strong><\/p>\n\n\n\n

In Sweden, AI-Ready operator, EcoDataCenter, has also used this form of energy tech and aimed to build a truly sustainable data centre<\/a> at the heart of Falun. As such, the company needed safe and reliable power systems that would ensure uptime and energy efficiency, with the opportunity for heat reuse to contribute towards delivering a low Power Usage Effectiveness rating (PUE).<\/p>\n\n\n\n

EcoDataCenter partnered with Schneider Electric from the outset, using its energy technology and digital software to support more than 3,750 kW of customer load, with connected sensor and meter monitoring for granular facility temperature control. The data centre achieves new levels of efficiency, with an EER of 132 and a PUE of 1.15 by reusing low-grade waste heat.<\/p>\n\n\n\n

These examples, alongside many others in\u00a0Clondalkin\u00a0and\u00a0Tallaght<\/a>\u00a0in Ireland,\u00a0Frankfurt<\/a>\u00a0in Germany,\u00a0Hamina\u00a0in Finland and\u00a0Odense<\/a>\u00a0in Denmark, all highlight the growing opportunity for operators to achieve net zero efforts while contributing to local communities.<\/p>\n\n\n\n

From waste to value: End-to-end heat reuse<\/strong><\/p>\n\n\n\n

From a broader perspective, cooling sits at the centre of any effective heat reuse strategy and in order to enable heat reuse, several energy technologies must work together to capitalise on this opportunity.<\/p>\n\n\n\n

Schneider Electric is helping deliver heat-recovery-ready infrastructure at every layer of the stack, from co-engineered NVIDIA GB200 reference designs<\/a> and Motivair direct liquid cooling units, to its portfolio of data centre solutions – <\/a>bringing power, cooling, management and security to aid IT deployments in all environments, while enabling organisations to optimise cooling performance and unlock opportunities for heat reuse.<\/p>\n\n\n\n

At the cooling level, InRow cooling<\/a>, hot aisle containment,<\/a> high-efficiency chilled water systems, Coolant Distribution Units (CDUs), together with intelligent environmental monitoring<\/a>, all help operators maintain highly thermal environments while supporting heat recovery initiatives.<\/p>\n\n\n\n

Power systems, including UPS, breakers and smart panels, all help to reduce electrical losses with less waste heat in electrical rooms, meaning more predictable heat\u2011capture from IT loads.<\/p>\n\n\n\n

Finally, prefabricated data centres <\/a>deliver adaptable, high-performance designs that meet complex infrastructure needs across diverse locations and environments. These scalable architectures enable organisations to increase compute density, integrate liquid cooling technologies<\/a>, expand heat recovery systems and connect with district heating networks – adapting to evolving sustainability requirements which are critical, as organisations modernise to support AI demand.<\/p>\n\n\n\n

Denser compute environments, for example, generate more concentrated heat, making reuse more effective when properly managed, which also depends on enhanced operational intelligence. With DCIM software, organisations can gain real-time visibility into power, cooling and environmental performance, while delivering controlled thermal conditions required for efficient heat recovery.<\/p>\n\n\n\n

Heat reuse at the intersection of demand, design and sustainability<\/strong><\/p>\n\n\n\n

As demand for greater capacity and the complexity of digital infrastructure grows, sustainability ambitions do not have to take a back-step. The defining characteristics of equipment that makes accelerated compute and AI possible are providing new opportunities to work meaningfully towards net zero, while presenting opportunities for operators to support and contribute to local communities.<\/p>\n","protected":false},"excerpt":{"rendered":"

Schneider Electric\u2019s Matthew Baynes, VP, Secure Power & Data Centres in the UK & Ireland, explores how turning waste energy from AI and high performance infrastructure can be a net zero advantage. There is an old bit of wisdom that asserts in every problem there is an opportunity. Accelerated compute and AI adoption are driving […]<\/p>\n","protected":false},"author":21,"featured_media":170824,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[21836,55,33,20673,36,7359,22293,13207,93],"tags":[158,11327,11337,6189,35,19743,4364,26069,774,16443],"class_list":["post-170823","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-data-centers","category-data-centres","category-energy","category-expert-opinion","category-green","category-industry-expert","category-innovation","category-thought-leadership","category-top-stories","tag-ai","tag-compute","tag-cooling","tag-data-centres-2","tag-energy","tag-heat-reuse","tag-infrastructure","tag-matthew-baynes","tag-schneider-electric","tag-thought-leadership"],"acf":[],"publishpress_future_workflow_manual_trigger":{"enabledWorkflows":[]},"_links":{"self":[{"href":"https:\/\/www.intelligentcio.com\/eu\/wp-json\/wp\/v2\/posts\/170823","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.intelligentcio.com\/eu\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.intelligentcio.com\/eu\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.intelligentcio.com\/eu\/wp-json\/wp\/v2\/users\/21"}],"replies":[{"embeddable":true,"href":"https:\/\/www.intelligentcio.com\/eu\/wp-json\/wp\/v2\/comments?post=170823"}],"version-history":[{"count":26,"href":"https:\/\/www.intelligentcio.com\/eu\/wp-json\/wp\/v2\/posts\/170823\/revisions"}],"predecessor-version":[{"id":171363,"href":"https:\/\/www.intelligentcio.com\/eu\/wp-json\/wp\/v2\/posts\/170823\/revisions\/171363"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.intelligentcio.com\/eu\/wp-json\/wp\/v2\/media\/170824"}],"wp:attachment":[{"href":"https:\/\/www.intelligentcio.com\/eu\/wp-json\/wp\/v2\/media?parent=170823"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.intelligentcio.com\/eu\/wp-json\/wp\/v2\/categories?post=170823"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.intelligentcio.com\/eu\/wp-json\/wp\/v2\/tags?post=170823"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}