Airsys urges UK MPs to focus on existing data centres

Airsys, a provider of data centre cooling systems, has called on MPs to consider upgrading existing data centre sites as part of the UK’s approach to expanding AI capacity. The cooling technology company will make the case for improving existing data centre assets in response to an inquiry launched by the All-Party Parliamentary Group (APPG) for Data Centres. Matthew Thompson, UK Managing Director at Airsys, says, “AI Growth Zones are clearly part of the mix, but these will take time to deliver. UK businesses need access to AI now, and better utilisation of existing sites has to be part of the answer.” The inquiry follows the publication of the APPG Insights report in May, which highlighted industry concerns around energy supply and grid access, energy costs, water use, planning, and sustainability. Matthew continues, “While government policy has emphasised construction of new data centres and AI Growth Zones, these will take years to come on stream. Planning and local community agreement is one hurdle, but getting sufficient power onto these sites is the biggest bottleneck for the industry. “The government policy is also heavily focused on regional developments, which has raised concerns amongst some operators about latency issues as inference becomes more important. “In the meantime, businesses need compute capacity now in strategic locations, that don’t compromise on latency.” Airsys is proposing that retrofitting existing sites and developing new capacity on brownfield sites should form part of the UK’s data centre and AI infrastructure agenda. Retrofitting could increase compute capacity Matthew notes, “The Government has correctly identified the UK’s massive potential for economic growth through data centres and AI. But, focusing purely on longer-term projects could mean more immediate opportunities to boost our AI capacity are lost.” Airsys argues that retrofitting legacy sites with liquid cooling could improve the use of available power by reducing cooling requirements and allowing more power to be allocated to computing. Data centre operators commonly use power usage effectiveness (PUE) to measure energy efficiency. Airsys has proposed two additional metrics - power compute effectiveness (PCE) and return on invested power (ROIP) - to assess the utilisation of data centre power and identify underused capacity. The company says deploying liquid cooling at existing sites, alongside developing capacity closer to users, could increase compute density and utilisation whilst reducing energy and water consumption. It also argues that a focus on longer-term regional AI Growth Zones, combined with fragmented local planning and policy, could make it more difficult to improve existing sites. Airsys will submit formal evidence to the inquiry, focusing on policies for edge environments and retrofitting alongside larger-scale data centre developments. Matthew concludes, “Clear, unified direction from central government and regulators is urgently required to ensure UK businesses get access to the capacity they need, where they need it. “We want to ensure the Government, legislators, and regulators are aware of all the tools available to ensure the country is able to fully realise the benefit of all its AI and data centre infrastructure." For more from Airsys, click here.

800 VDC arc flash risk assessed in new Schneider study

Global energy technology company Schneider Electric has published an analysis of arc flash risk in 800 VDC power architectures, examining how different system designs, capacitor placement, and fault-clearing behaviour affect the potential for electrical hazards. The analysis uses deployment scenarios based on hyperscaler design patterns and compares two emerging 800 VDC architectures with different configurations. It finds that, even under conservative assumptions where capacitors dominate fault behaviour, arc flash risk in 800 VDC systems can be managed and, in some cases, is comparable with typical AC systems. The study also finds that advanced software and digital twins can provide more detailed modelling of arc flash risk than simplified calculation methods. The research comes as 800 VDC architectures are being considered for higher-density AI data centre racks. NVIDIA and other technology companies, including Schneider Electric, are working on power infrastructure intended to support racks rated at 400 kW and above. As power densities increase, 800 VDC distribution is being considered for supplying higher-power racks while limiting distribution losses. However, higher operating voltages also require greater understanding of fault behaviour, protection coordination, and safe working practices. Manish Kumar, EVP Secure Power & Data Centers, Schneider Electric, says, "800 VDC power distribution represents a significant shift in data centre design, but it also introduces safety considerations that need to be studied extensively. "Our work with some of the world’s leading hyperscalers provides engineers and safety professionals with one of the first practical frameworks for evaluating arc flash risks, providing a structured approach to understanding fault behaviour, establishing safe work practices, and designing effective protection schemes. "Our goal is to help the industry move towards higher-voltage architectures with confidence and safety." Two 800 VDC architectures examined The analysis assesses rack-level and facility-level 800 VDC architectures, representing two implementation approaches being considered for data centre power distribution. It compares standards-based methods, transient simulation, and system-level modelling, finding that existing arc flash frameworks can be applied to 800 VDC systems when the architecture and time-dependent behaviour of faults are taken into account. For rack-level 800 VDC architectures, a sidecar (also known as a power rack) case study used conservative methods and assumptions. The results showed incident energy below the referenced 1.2cal/cm² PPE threshold, even without protection devices. The facility-level case study showed the potential for slightly higher incident energy than the rack-level design. This assessment also used a conservative architecture without overcurrent protection. The analysis examined system topology and the effect of fault locations upstream and downstream of reverse-blocking diodes on back-feed, peak current, and arc flash outcomes. When fault contribution is limited in time using standard protection devices, the study found that arc flash energy can be reduced to levels appropriate for the working environment and generally comparable with common AC architectures. The research identifies several factors that influence arc flash risk in 800 VDC systems: • Transient behaviour matters — Arc flash is driven by time-dependent fault currents, with capacitor discharge dominating the first milliseconds of an event.• Simulation can improve accuracy — Transient simulation and power system analysis tools can provide a more detailed assessment than simplified DC arc flash calculations, which the study finds can overestimate risk in capacitor-dominated systems.• System design affects risk — Capacitor placement, reverse-blocking devices, and millisecond-scale protection can influence arc flash outcomes. Tanuj Khandelwal, CEO of ETAP, notes, "Industry standards remain essential for arc flash and electrical safety, but traditional methods can be overly conservative because they do not fully reflect how complex DC systems operate. "To understand real risk, engineers must evaluate system topology, fault behaviour, protection coordination, converter response, switching logic, and active protection schemes. "ETAP enables teams to model and validate 800V DC systems as they perform, helping move from conservative assumptions to more accurate, AI-augmented, physics-based safety and operational decisions." The study concludes that system architecture and protection strategies are significant factors in determining arc flash risk in 800 VDC systems. It finds that overall risk can remain low and, in some cases, comparable with typical AC distribution, including when standard time-based protection devices are used. Schneider Electric says the work builds on its previous arc flash safety testing and forms part of its work on 800 VDC power architectures for higher-density rack systems. The company has also conducted testing of live-swap power capabilities in 800 VDC systems for maintenance applications. The full findings have been published in the new whitepaper, titled DC Arc Flash Analysis: A Practical Study on 800 VDC in Data Centers. For more from Schneider Electric, click here.

AVK secures $1bn investment from Partners Group

AVK, a provider of power systems and electrical infrastructure for data centres, has secured an initial investment of more than $1 billion (£743 million) from global private markets investment firm Partners Group, which will take a majority stake in the UK-based power infrastructure company. The investment is intended to support AVK's expansion and fund the development of on-site power infrastructure for data centre operators under an 'energy as a service' (EaaS) model. Ben Pritchard will retain a significant shareholding and continue as Chief Executive Officer, alongside the existing leadership team. The investment is the first external funding round in AVK's 36-year history, with it continuing to supply prime, standby, and dispatchable power infrastructure for data centres and other critical applications. AVK says the funding will support its strategy of funding, developing, owning, and operating on-site power infrastructure, including microgrids, whilst the company currently has a pipeline of more than 2GW. Under the EaaS model, data centre operators can procure energy through power purchase agreements (PPAs), rather than directly funding and managing large-scale on-site energy developments. AVK's existing operations are supported by its manufacturing facility in Haydock, in the North West of England, and a workforce of nearly 400 people across 10 hubs in the UK and Europe. Ben Pritchard, Chief Executive Officer at AVK, comments, “Speed-to-power is now a defining opportunity for European data centre operators. Our new partnership with Partners Group will allow us to meet our customers exactly where the market demands. "From the moment we launched our first microgrid, we recognised the challenge and the opportunity facing developers and operators globally. By adding capital to our power solutions portfolio, we can turn speed-to-power from an ambition into action. "I am excited to lead AVK into this new chapter alongside Partners Group, leveraging the firm's deep operational expertise in the data centre sector and power markets.” Investment targets on-site power infrastructure Partners Group has already invested in decentralised energy and data centres in Europe, including the pan-Nordic data centre platform atNorth. It also has experience investing in behind-the-meter energy infrastructure for data centres in the USA. Nicholas Pepper, Managing Director, Infrastructure at Partners Group, notes, "AI is driving one of the largest infrastructure buildouts in decades, and access to power is becoming a defining constraint. This constraint and lengthening connection queues are critical bottlenecks to growth in the European data centre market, which on-site generation can alleviate by accelerating speed-to-power. "AVK, with its deep expertise, track record, and pan-European footprint, is well positioned to address this issue as a one-stop-shop for data centre power solutions. We see an exciting growth opportunity for AVK and we look forward to supporting the management team in its next chapter." Earlier in 2026, AVK energised a large-scale data centre microgrid at Pure DC's campus in Dublin. The company says the project was the first of its scale for a data centre microgrid in Europe. For more from AVK, click here.

Vultr scales AI infrastructure with new AMD GPUs

Vultr, a privately held cloud infrastructure company, has announced plans to offer the AMD Instinct MI455X GPU, alongside support for the AMD Helios rack-scale architecture, with pre-orders now open ahead of availability in Q4. The cloud infrastructure provider says the platform is designed for AI training, inference, fine tuning, agentic AI, and high-performance computing (HPC). According to the company, the infrastructure is intended to support organisations moving AI workloads from development into production, particularly those requiring high-memory, rack-scale computing. J.J. Kardwell, CEO of Vultr, comments, "Customers are rapidly moving from AI experimentation into production and, increasingly, that means agentic and inference-heavy workloads that require infrastructure built for scale. "The AMD Instinct MI455X and Helios rack-scale architectures give our customers the flexibility, control, and price-to-performance they need to scale their AI initiatives. Making this architecture available across Vultr's global infrastructure enables teams to build and deploy without constraints." AI infrastructure targets large-scale workloads The AMD Instinct MI455X GPU is based on AMD's CDNA architecture and features HBM4 memory, increased memory bandwidth, and support for low-precision AI data types. The Helios rack-scale architecture supports up to 72 GPUs in a single deployment. Vultr says the platform will also support direct liquid cooling to improve thermal management and energy efficiency in higher-density deployments. The infrastructure is complemented by AMD's ROCm software platform and Pensando networking technology, which are designed to improve AI training, inference performance, and scalability. IDC forecasts a tenfold increase in the number and complexity of enterprise AI agents over the next five years, increasing demand for computing infrastructure capable of supporting repeated inference workloads. Dave McCarthy, Group Vice President, Cloud and Datacenter Infrastructure at IDC, notes, "As enterprises transition into the agentic era, the massive influx of complex, data-intensive workflows requires a fundamental shift in infrastructure design. "Vultr's deployment of the AMD Instinct MI455X GPU provides the massive memory capacity and rack-scale efficiency necessary to support the continuous inference loops that define next-generation AI workloads." Vultr says reserved capacity for deployments based on the AMD Instinct MI455X GPU and AMD Helios architectures is now available for pre-order, with deployments planned from 2027 through to 2028. For more from Vultr, click here.

Sharon AI signs $373m cloud agreement

Sharon AI, an Australian neocloud provider, has signed a five-year cloud computing agreement with a global artificial intelligence platform worth $373 million (£277 million). Under the agreement, Sharon AI expects to deploy cloud computing infrastructure across its AI facilities in Australia, with revenue expected to begin in the first quarter of 2027. The company says its total AI factory capacity remains at 132MW, with 120MW now contracted to end customers. It also plans to increase the number of NVIDIA GPUs across its AI factory platform from 62,000 to 64,000 by mid-2027. The initial deployment under the agreement is expected to use 2,048 NVIDIA Blackwell Ultra B300 GPUs. Further deployments may take place during the five-year agreement, subject to customer requirements and the contract terms. An expansion of Australian AI infrastructure capacity Sharon AI says the agreement will contribute to the expansion of AI compute capacity in Australia and support the development of domestic AI infrastructure. James Manning, co-founder and Chief Executive Officer of Sharon AI, comments, "This agreement represents an important milestone in the continued expansion of Sharon AI's customer base and contracted AI infrastructure capacity. "As organisations increasingly seek access to sovereign, high-performance AI compute, we remain focused on delivering scalable infrastructure that supports the evolving needs of AI platforms, enterprises, and governments. "We continue to see strong demand for AI infrastructure in Australia and across the Asia-Pacific region, and this agreement reflects our strategy of securing long-term customer commitments as we expand our AI factory platform." The company has said the agreement is also expected to support investment in Australian digital infrastructure and provide additional opportunities for businesses, researchers, and government organisations to access AI compute capacity domestically.

Cabling Group named Panduit's European Platinum Partner

The Cabling Group has been awarded Platinum Partner status by Panduit, a manufacturer of electrical and network infrastructure hardware, making it the company's only current Platinum Partner in Europe. Panduit's Platinum Partner designation is its highest level of channel recognition and is awarded to organisations based on technical expertise, customer service, and their relationship with the company. The Cabling Group is a UK specialist in network infrastructure and connectivity. The partnership brings together Panduit's infrastructure portfolio and The Cabling Group's technical expertise and customer relationships. Jens Holzhammer (pictured above), Managing Director EMEA at Panduit, comments, "This partnership reflects our shared commitment to innovation, operational excellence, and delivering long-term value to customers. "By combining Panduit's industry-leading infrastructure solutions with The Cabling Group's technical expertise and market reach, we are well positioned to accelerate growth across the network infrastructure business. We are very proud to have them as a Platinum Partner." Graham Smith, Managing Director of The Cabling Group, adds, "We're delighted to have been awarded Panduit Platinum Partner status. This is a fantastic achievement that reflects the hard work, technical expertise, and commitment of our team. "Achieving Platinum status strengthens our partnership with Panduit and enhances our ability to continue delivering high-quality infrastructure solutions for our customers." Partnership expands network infrastructure focus Michael Kemmerling, Business Director Central Europe North at Panduit, notes, "Customers today are looking for trusted partners that can help simplify complexity while supporting scalability, reliability, and performance. "The Cabling Group brings deep expertise and a strong reputation in the market, and we are proud to strengthen our collaboration through this Platinum Partner designation." The companies say the partnership will support customers investing in network infrastructure, with a focus on scalability, reliability, and performance across the UK and wider European market. For more from Panduit, click here.

Wolong's PMDD motors target data centre cooling

Wolong Electric America, a developer of industrial motor and drive technology, has highlighted its permanent magnet direct drive (PMDD) motors for data centre and industrial cooling applications. The motors are manufactured in Monterrey, Mexico, and use a 48-pole design intended to provide a compact alternative to conventional induction motors. The PMDD design removes the need for belts and sheaves by connecting the motor directly to the fan. According to Wolong, this reduces mechanical complexity, maintenance requirements, and potential points of failure. The motors use a rare earth magnet core, which the company says generates stronger magnetic fields than ferrite-based alternatives while allowing for a smaller footprint. A variable frequency drive (VFD) controls the motors, providing variable speed operation and controlled acceleration and deceleration. The motors have a 4:1 turndown ratio, allowing them to maintain torque at lower speeds for applications where cooling demand varies. Direct drive design reduces mechanical load The direct coupling between the motor and fan reduces radial load on bearings and associated mechanical stress by removing the belt drive. Wolong states that the motors are approximately 20% more efficient than conventional induction motors. The design also provides smooth starts without inrush current, which can reduce electrical stress on connected equipment. The motors are intended for use in data centres, refineries, heat exchangers, cooling systems, and other OEM-designed equipment. Wolong's Monterrey manufacturing operation also provides regional manufacturing, service, and customer support for North American customers. The company says the motors are designed with anticipated US Department of Energy efficiency requirements in mind and can be used when upgrading existing equipment or developing new cooling systems. For more from Wolong, click here.

Yondr acquires site for Northern Virginia data centre

Yondr Group, a global developer, owner, and operator of hyperscale data centres, has acquired a 40-acre (16.2-hectare) site in Manassas, Virginia, USA - in partnership with funds and accounts managed by Cerberus Capital Management and its affiliates - which is expected to support the construction of a 72MW data centre campus, with operations planned to begin in 2029. The project is intended to provide additional capacity for hyperscale customers in Northern Virginia, supporting applications including cloud, enterprise, and artificial intelligence workloads. Northern Virginia is already an established data centre market, with a significant presence of hyperscale and cloud operators, extensive fibre connectivity, and connections to major hubs across the northeastern United States. Aaron Wangenheim, CEO of Yondr, comments, "We continue to see strong demand for well-located capacity across our global portfolio, including Northern Virginia, and this acquisition marks another important step in growing our North American footprint. "Bringing a project of this scale online in a market like Northern Virginia takes deep operational expertise and strategic, sophisticated capital, and our partnership with Cerberus brings both together to deliver the capacity hyperscale customers increasingly need." 72MW campus planned for 2029 The project is expected to have power available in the near term, with a target ready-for-service date of 2029. Tom Wagner, Senior Managing Director and Head of North American Real Estate at Cerberus, says, "We are pleased to partner with Yondr to deliver a high-quality project in Northern Virginia that is well positioned to support hyperscale demand. "With near-term power availability in a historically constrained market and a 2029 ready-for-service date, this project will be well positioned to support continued customer demand while creating long-term value for our partners and investors. "We look forward to advancing this project alongside Yondr and identifying compelling opportunities to invest in high-quality real assets supported by the strong fundamentals in the digital economy." For more from Yondr, click here.

FTTH Council Europe becomes Fibre Council Europe

The FTTH Council Europe, an industry association promoting fibre connectivity across Europe, has rebranded as Fibre Council Europe, marking an expansion of its remit to represent the wider fibre infrastructure ecosystem across Europe. The organisation says the new name reflects its broader focus beyond fibre-to-the-home (FTTH), encompassing fixed connectivity, the fibre infrastructure supporting data centres, mobile networks, and the digital services used by businesses, public institutions, and consumers. The rebrand includes a new visual identity and the tagline: "Powering Europe's digital future". Stefano Fogli, President of Fibre Council Europe, says, "Our new name recognises what our members have known for years: fibre is the essential infrastructure of the present and the future-proof foundation of what comes next, with the potential to evolve to ever-higher speeds. "As Fibre Council Europe, we will continue to bring together the entire fibre community to accelerate deployment, drive take-up, and ensure Europe's digital competitiveness. Our mission expands, and our name now matches its full scope." Conference renamed as organisation broadens remit The Council's annual conference has also been renamed Fibre Horizons. The event, which the organisation describes as Europe's largest gathering of the fibre community, will take place on 17–18 March in Milan, Italy, bringing together operators, vendors, investors, and policymakers. According to the Council, its governance, membership, and core activities will remain unchanged, while its work will expand into areas including data centres and the transition from hybrid fibre-coaxial (HFC) networks to FTTH. Vincent Garnier, Director General of Fibre Council Europe, comments, "This is an evolution, not a departure. Everything our members and partners value about the Council - our advocacy, our research, our community - carries forward under the new name and the expanded scope. What changes is that our identity now reflects the breadth of the fibre ecosystem we represent." The organisation says the transition to the new branding will be phased in over the coming weeks, with its previous branding, email addresses, and web addresses remaining in use during the changeover period. For more from Fibre Council Europe, click here.

Reliable liquid cooling infrastructure with aquatherm piping

Rising rack densities, AI workloads, and 24/7 availability place new demands on data centre cooling. Conventional air-based concepts increasingly reach their limits when it comes to removing high heat loads efficiently and reliably. aquatherm blue is a PP-RCT piping system designed specifically for liquid cooling circuits in data centres and similar mission-critical environments. It provides the backbone for closed chilled-water loops, direct-to-rack cooling, and integration with heat exchangers or free cooling systems. The system’s corrosion-free material and smooth inner surface support stable flow rates over the long term, helping maintain performance and reduce unplanned interventions. Lightweight components and heat-fusion jointing simplify installation in raised floors, technical rooms, and retrofits. With flexible layouts and a wide range of dimensions and fittings, aquatherm blue enables planners and operators to adapt cooling infrastructure to current and future IT loads. For more information, click here.